1. Publication
- Title: Recombinant Factor VIIa versus Placebo for Spontaneous Intracerebral Haemorrhage within 2 h of Symptom Onset
- Acronym: FASTEST
- Year & Journal: The Lancet, published February 4, 2026 (2026;407(10530):773-783)
- Citation: Broderick JP, Naidech AM, Elm JJ, et al. Recombinant factor VIIa versus placebo for spontaneous intracerebral haemorrhage within 2 h of symptom onset (FASTEST): a multicentre, double-blind, randomised, placebo-controlled, phase 3 trial. Lancet. 2026;407(10530):773-783. doi:10.1016/S0140-6736(26)00097-8
2. Context & Rationale
Background: ICH accounts for ~15% of strokes but a disproportionate ~50% of stroke deaths. Most hematoma expansion occurs within the first 2–3 hours after onset. Recombinant factor VIIa (rFVIIa) had previously been shown to slow bleeding in ICH, but no hemostatic agent had ever been shown to improve clinical outcomes — earlier trials may not have treated patients quickly enough or selected the right population.
Research Question/Hypothesis: In patients with acute spontaneous ICH treated within 2 hours of onset, does rFVIIa improve functional outcome (modified Rankin Scale) at 180 days compared with placebo, while limiting hematoma/IVH growth?
Why This Matters: FASTEST was designed to test hemostatic therapy in the population and time window most likely to benefit — excluding very elderly patients and those with extremely large hemorrhages unlikely to have good outcomes regardless of bleeding control — addressing whether stopping hematoma growth translates into better survival/function.
3. Design & Methods
- Study Type: Multicenter, double-blind, randomized, placebo-controlled, adaptive, phase 3 trial
- Setting & Centers: 93 sites, USA, Japan, Canada, Spain, Germany, UK (largest hyperacute stroke trial to require brain imaging before treatment)
- Population:
- Inclusion: Adults 18–80y, spontaneous ICH 2–60 mL, IVH <2/3 of one lateral ventricle or <1/3 of both, GCS ≥8, no recent ischemic stroke/MI, no recent anticoagulation, treated within 2h of onset/last known well
- Exclusions: Recent ischemic stroke or MI, recent anticoagulant use, other structural cause of ICH
- Intervention: rFVIIa 80 μg/kg IV over 2 minutes
- Comparator: Identical-appearing placebo
- Blinding: Double-blind (participants and investigators)
- Statistical Power & Follow-Up: Planned 860 patients; primary efficacy outcome mRS at 180 days; primary safety outcome life-threatening thromboembolic events within first 4 days. Trial stopped early for futility at second pre-planned interim analysis after 626 of planned 860 randomized (434 with primary-outcome data available).
4. Key Results
Stopped early for futility. 3288 screened (Dec 3, 2021 – Oct 1, 2025); 626 randomized and included in ITT analysis (328 [52%] rFVIIa, 298 [48%] placebo).
Outcome | rFVIIa | Placebo | Effect Size | 95% CI | p-value | Clinical Notes |
Functional outcome, mRS at 180 days (primary efficacy) | — | — | Adjusted common OR 1.09 | 0.79–1.51 | 0.61 | No significant functional benefit |
Life-threatening thromboembolic events, first 4 days (primary safety) | 15/328 (<5%) | 4/298 (1%) | RR 3.41 | 1.14–10.15 | 0.020 | Significant excess with rFVIIa |
Hematoma/IVH growth | Reduced with rFVIIa | — | Not reported as single figure | Not reported | — | rFVIIa significantly reduced bleeding progression (secondary outcome) |
5. Internal Validity Assessment
- Randomization & Allocation: 1:1 randomization; double-blind with identical-appearing placebo — strong allocation concealment and blinding.
- Protocol Adherence & Separation: Rapid, protocolized dosing (within 2h of onset) achieved as designed; hematoma growth reduction confirms biological activity/on-target effect.
- Blinding & Detection Bias: Full double-blind design (participants and investigators) — minimizes both performance and detection bias.
- Missing Data & Sensitivity Analyses: Trial stopped early at interim analysis (626/860 planned, 434 with primary-outcome data available at the point of the futility decision) — early stopping for futility is itself a form of "missing data by design" relative to the full planned sample.
- Overall Internal Validity Conclusion: Strong for the null functional-outcome finding (double-blind, adequately event-rich for the interim futility determination) and for the safety signal (significant, despite early stopping) — though early termination means the trial cannot fully exclude a smaller functional benefit than the effect size the total sample was originally powered to detect.
6. External Validity Assessment
- Population Representativeness: Carefully selected — ICH volume 2–60 mL, GCS ≥8, age 18–80y, minimal IVH, treated within 2h — explicitly designed to capture patients most likely to benefit, meaning results do not generalize to very elderly patients, larger hemorrhages, or those presenting later.
- Practice Context: Requires very rapid imaging and treatment initiation (within 2h) across a large international network including mobile stroke units — achievable in well-resourced hyperacute stroke systems, but a significant infrastructure requirement.
- Overall External Validity Conclusion: Good for the specific, carefully selected early-presenting ICH population and health systems capable of achieving very rapid (<2h) treatment; not generalizable to broader, less-selected, or later-presenting ICH populations by design.
7. Strengths & Limitations
Strengths:
- Largest hyperacute stroke trial to date requiring brain imaging before treatment
- Rigorous double-blind, placebo-controlled, adaptive design with international reach (93 sites, 6 countries)
- Successfully demonstrated on-target biological effect (reduced hematoma/IVH growth), isolating the question of whether stopping bleeding translates to functional benefit
- Confirms and extends prior mechanistic knowledge (rFVIIa slows bleeding) with a large, well-designed efficacy/safety trial
Limitations:
- Stopped early for futility — did not reach the planned sample size (626/860)
- Significant excess of life-threatening thromboembolic events with rFVIIa (RR 3.41)
- Narrow, carefully selected population limits generalizability
- Despite the population selection specifically designed to maximize chance of benefit, no functional benefit was seen — a genuinely important negative finding for the field
8. Interpretation & Practice Impact
- Clinical Implications: Does not support use of rFVIIa for spontaneous ICH, even in a population and time window specifically selected to maximize the chance of benefit — the biological effect (reduced bleeding) did not translate into improved functional outcomes, and came with a significant increase in life-threatening thromboembolic complications.
- Mechanistic Coherence: This is the clearest demonstration yet that reducing hematoma growth alone is not sufficient to improve functional recovery in ICH — an important conceptual result for the field, given decades of hypothesis that hemostatic therapy "should" work if it stops bleeding.
- Systems-Level Takeaway: Further research on hemostatic therapy is being directed toward identifying a more narrowly defined subset of patients at highest risk of continued bleeding, rather than broader ICH populations.
9. Controversies & Subsequent Evidence
- Editorial Commentary: An accompanying Lancet editorial ("Testing early haemostatic therapy for acute intracerebral haemorrhage") explicitly cautions against over-interpreting subgroup signals and emphasizes that demonstrating an effect on a plausible mechanistic surrogate (hematoma growth) does not guarantee a functional-outcome benefit — a lesson FASTEST reinforces empirically.
- Guideline Integration: Consistent with (and reinforces) the established pattern that no hemostatic agent has yet been shown to improve clinical outcomes in unselected or broadly-selected ICH; further testing of rFVIIa in patients at the very greatest risk of continued bleeding is described as ongoing.
10. Summary & Executive Takeaway
Summary: FASTEST, stopped early for futility after 626 of a planned 860 patients, randomized adults with early-presenting (within 2h), carefully selected spontaneous ICH to rFVIIa or placebo. rFVIIa reduced hematoma/IVH growth but showed no significant functional benefit at 180 days (adjusted common OR 1.09, 95% CI 0.79–1.51, P=0.61) and significantly increased life-threatening thromboembolic events (RR 3.41, 95% CI 1.14–10.15, P=0.020).
Overall Takeaway: Even in a population and ultra-early time window specifically selected to maximize benefit, rFVIIa's proven ability to slow ICH bleeding did not translate into improved functional recovery, while significantly increasing thromboembolic risk — a genuinely important negative result reinforcing that hematoma growth is a surrogate, not a guaranteed proxy, for patient-centered outcomes in ICH.
11. Bibliography
- Mayer SA, Brun NC, Begtrup K, et al. Recombinant activated factor VII for acute intracerebral hemorrhage. N Engl J Med. 2008;358(20):2127-2137.
- Al-Shahi Salman R. Testing early haemostatic therapy for acute intracerebral haemorrhage [editorial]. Lancet. 2026;407(10530):735-737.