TL;DR: 🟡 Genuinely mixed — andexanet gave superior hematoma-expansion control in factor Xa-associated ICH, but significantly increased thrombotic events (10.3% vs 5.6%) including ischemic stroke, with no net clinical outcome benefit.
1. Publication
- Title: Andexanet for Factor Xa Inhibitor–Associated Acute Intracerebral Hemorrhage
- Acronym: ANNEXA-I
- Year & Journal: New England Journal of Medicine, published May 16, 2024 (390(19):1745-1755)
- Citation: Connolly SJ, Sharma M, Cohen AT, et al; ANNEXA-I Investigators. N Engl J Med. 2024;390(19):1745-1755. doi:10.1056/NEJMoa2313040
2. Context & Rationale
Background: Factor Xa inhibitor-associated ICH requires rapid reversal; hematoma expansion predicts poor outcomes. Andexanet alfa specifically sequesters factor Xa inhibitors, restoring hemostasis. The prior single-arm ANNEXA-4 cohort study suggested ~80% hemostatic efficacy but ~10% thrombotic events, without a comparator arm.
Research Question/Hypothesis: In factor Xa inhibitor-associated ICH, does andexanet alfa achieve better hemostatic efficacy than usual care (predominantly 4-factor PCC)?
3. Design & Methods
- Study Type: Randomized, controlled, phase 3/4 trial with blinded endpoint adjudication
- Population: 530 patients with ICH who took factor Xa inhibitors within 15h of randomization (protocol amendment mid-enrollment restricted to intracerebral, excluding subdural/subarachnoid)
- Intervention: Andexanet alfa
- Comparator: Usual care (predominantly 4-factor PCC)
- Blinding: Open-label treatment; blinded end-point adjudication committee and core imaging laboratory
- Statistical Power & Follow-Up: Primary: hemostatic efficacy (hematoma expansion ≤35% at 12h, NIHSS increase <7 points, no rescue therapy 3-12h). Sample size calculated for 90% power to detect 10-percentage-point difference.
4. Key Results
530 patients randomized.
Outcome | Andexanet Alfa | Usual Care | Effect Size | p-value | Notes |
Hemostatic efficacy (primary) | Higher proportion | Lower proportion | Not reported as single ratio in accessible text | — | Significantly better hematoma-expansion control with andexanet |
Thrombotic events | 10.3% | 5.6% | Diff 4.6pp | 0.048 | Significantly increased with andexanet |
Ischemic stroke | 6.5% (17 patients) | 1.5% (4 patients) | — | — | Notable component of thrombotic events |
Functional outcome (mRS), 30-day death | No appreciable difference | No appreciable difference | — | — | No clinical outcome benefit despite hemostatic efficacy |
5. Internal Validity Assessment
Randomized, controlled trial with blinded endpoint adjudication and a blinded core imaging laboratory for hematoma volume assessment. Overall: Strong — rigorous, blinded ascertainment of both the primary hemostatic endpoint and thrombotic safety events; the significant hemostatic benefit alongside a significant safety signal (more thrombotic events) provides a clear, if genuinely mixed, result.
6. External Validity Assessment
International population of factor Xa inhibitor-associated ICH patients — broadly relevant to the growing population on DOACs; mid-trial protocol amendment restricting to intracerebral (excluding subdural/SAH) narrows the population somewhat.
7. Strengths & Limitations
Strengths: Blinded endpoint adjudication and imaging core lab; addresses a genuine, increasingly common clinical scenario (DOAC-associated ICH); demonstrates clear pharmacological on-target effect (rapid anti-factor Xa activity reduction).
Limitations: No clinical outcome benefit (mRS, 30-day mortality) despite better hemostatic control; significantly increased thrombotic events including ischemic stroke; industry funding (AstraZeneca/Alexion); an accompanying editorial specifically noted the timing/severity of ischemic strokes was not well described, and questioned whether atrial fibrillation status was adequately characterized in affected patients.
8. Interpretation & Practice Impact
Supports andexanet's superior hemostatic efficacy for factor Xa inhibitor-associated ICH, but the significantly increased thrombotic event rate (including ischemic stroke) creates a genuine net-benefit uncertainty — the authors themselves acknowledge "determining the potential net benefit... is challenging."
9. Controversies & Subsequent Evidence
An accompanying NEJM editorial (Wade Smith, Claude Hemphill) explicitly raised concerns about incomplete characterization of the ischemic strokes' timing/severity and atrial fibrillation status in affected patients — a genuine, unresolved interpretive question. Given the mixed efficacy/safety picture and cost, many centers have continued using 4-factor PCC for DOAC reversal, per independent commentary (SGEM), reflecting real-world hesitancy to adopt andexanet despite its regulatory approval and hemostatic superiority.
10. Summary & Executive Takeaway
Summary: ANNEXA-I randomized 530 patients with factor Xa inhibitor-associated ICH to andexanet alfa or usual care (mostly 4-factor PCC). Andexanet achieved significantly better hemostatic efficacy, but thrombotic events were significantly more common (10.3% vs 5.6%, P=0.048), including ischemic stroke (6.5% vs 1.5%); functional outcomes and 30-day mortality showed no appreciable difference.
Overall Takeaway: Andexanet alfa provides genuinely superior hematoma-expansion control in factor Xa inhibitor-associated ICH, but this comes with a real, statistically significant increase in thrombotic events including ischemic stroke, and no demonstrated net clinical outcome benefit — a mixed result that has led many centers to continue favoring 4-factor PCC, reflecting genuine, unresolved uncertainty about andexanet's true net clinical value in this setting.
11. Bibliography
- Connolly SJ, Milling TJ, Eikelboom JW, et al. Andexanet alfa for acute major bleeding (ANNEXA-4). N Engl J Med. 2016;375:1131-1141.
- Smith WS, Hemphill JC. [Editorial]. N Engl J Med. 2024;390(19).