← Back
NeuroTrials.ai
Neurology Clinical Trial Database

FAST

Efficacy and Safety of Recombinant Activated Factor VII for Acute Intracerebral Hemorrhage

Year of Publication: 2008

Authors: Mayer SA, Brun NC, Begtrup K, ..., for the FAST Trial Investigators

Journal: New England Journal of Medicine

Citation: N Engl J Med 2008;358(20):2127-2137. DOI: 10.1056/NEJMoa0707534

Link: https://doi.org/10.1056/NEJMoa0707534

Bottom Line

In this phase 3 trial (FAST), hemostatic therapy with rFVIIa within 4 hours of ICH reduced 24-hour hematoma growth (11% vs 26% with 80 μg/kg) but did NOT reduce 90-day death or severe disability, and 80 μg/kg increased arterial thromboembolic events (9% vs 4%, P=0.04).

Major Points

  • Primary outcome (mRS 5 or 6 at 90 days) was not different across groups: 24% placebo, 26% rFVIIa 20 μg/kg, 29% rFVIIa 80 μg/kg — clearly negative for clinical benefit.
  • Hemostatic effect confirmed: mean 24-h ICH volume increase 26% (placebo), 18% (20 μg/kg), 11% (80 μg/kg, P<0.001 vs placebo); absolute reduction 3.8 mL with 80 μg/kg (95% CI 0.9-6.7, P=0.009).
  • Mortality at 90 days ~20% across all three arms; no survival benefit from rFVIIa.
  • Arterial thromboembolic serious adverse events were increased with 80 μg/kg vs placebo (9% vs 4%, P=0.04); troponin elevation 22% vs 15%; CT-defined cerebral infarction 4.7% vs 2.2%.
  • Randomization imbalances (more intraventricular hemorrhage, coma, and left ventricular hypertrophy in 80 μg/kg arm) and unexpectedly good placebo-group outcomes (mortality 19%, poor outcome 24%) may partly explain the discordance with the earlier phase 2b trial.
  • Post-hoc subgroup (age ≤70, ICH ≤60 mL, IVH ≤5 mL, treated ≤2.5 h; 19% of cohort): adjusted OR for poor 90-day outcome 0.28 (95% CI 0.08-1.06, P=0.03) — hypothesis-generating and motivated later spot-sign / early-treatment trials.
  • Findings largely ended routine consideration of rFVIIa for unselected ICH; interest continued only in ultra-early / spot-sign-positive subgroups (later tested by SPOTLIGHT, STOP-IT, TRAIGE, and FASTEST).

Design

Study Type: Phase 3 Randomized Controlled Trial

Randomization: 1

Blinding: Double-blind, placebo-controlled (block randomization by site)

Enrollment Period: May 2005 - February 2007

Follow-up Duration: 90 days

Centers: 122

Countries: Multinational (22 countries; sites in North America, Europe, Australia, Asia, South America, and the Middle East)

Sample Size: 841

Analysis: Intention-to-treat; primary outcome analyzed by logistic regression adjusted for age, sex, baseline ICH volume, prestroke mRS, and location (supratentorial vs infratentorial); last observation carried forward for surviving patients with missing data


Inclusion Criteria

  • Age 18 years or older
  • Spontaneous intracerebral hemorrhage documented by CT scan
  • CT performed within 3 hours after symptom onset
  • Able to receive study drug within 4 hours after onset of symptoms (and within 1 hour after baseline CT)
  • Written informed consent from patient or legally acceptable surrogate

Exclusion Criteria

  • Glasgow Coma Scale score of 5 or less
  • Surgical evacuation of hematoma planned within 24 hours
  • Secondary intracerebral hemorrhage from trauma, arteriovenous malformation, aneurysm, tumor, or other identifiable structural cause
  • Known use of oral anticoagulant therapy, thrombocytopenia, or coagulopathy
  • Acute sepsis, crush injury, or disseminated intravascular coagulation
  • Pregnancy
  • Previous disability (prestroke modified Rankin scale score >2)
  • Known thromboembolic disease within 30 days before enrollment (angina, claudication, deep-vein thrombosis, cerebral infarction, or myocardial infarction)

Baseline Characteristics

CharacteristicOverall Cohort (N=841)Placebo (n=268)rFVIIa 20 μg/kg (n=276)rFVIIa 80 μg/kg (n=297)
Mean Age (years)65 (range 23-97)
Sex - Male62%
Race - White69%
Race - Asian19%
Race - Black9%
Median Glasgow Coma Scale14 (range 6-15)
Mean NIHSS13 (range 0-38)
ICH Location - Deep gray matter78%
ICH Location - Lobar22%
Mean Baseline ICH Volume (mL)23.2 (range 0.3-153.4)
Time onset to baseline CT (min, mean ± SD)109 ± 39
Time baseline CT to treatment (min, mean ± SD)51 ± 17
Time onset to treatment (min, mean ± SD)160 ± 37
Treated within 2 h of onset17%
Treated within 3 h of onset72%
Intraventricular hemorrhage at baseline29%~35% (intermediate)41%
NoteBaseline characteristics broadly balanced; see randomization-imbalance notes below.Similar demographics to placebo group.Prognostically worse profile than placebo arm due to imbalance.
Left ventricular hypertrophy (baseline ECG)Higher than placebo
Coma at baseline (GCS 6-8)More frequent than placebo
Baseline total lesion volume (ICH+IVH+edema)Larger than placebo

Arms

FieldControlrFVIIa 20 μg/kgrFVIIa 80 μg/kg
InterventionSingle IV dose of matching placebo (reconstituted freeze-dried powder) over 2 minutes, given within 4 hours of ICH onsetRecombinant activated factor VII (NovoSeven, Novo Nordisk) 20 μg/kg IV over 2 minutes, single dose within 4 hours of onsetRecombinant activated factor VII (NovoSeven, Novo Nordisk) 80 μg/kg IV over 2 minutes, single dose within 4 hours of onset
DurationSingle doseSingle doseSingle dose

Outcomes

OutcomeTypeControlInterventionHR / OR / RRP-value
Poor outcome — severe disability or death (modified Rankin scale 5 or 6) at 90 daysPrimaryPlacebo: 24% · rFVIIa 20 μg/kg: 26% · rFVIIa 80 μg/kg: 29% · Not significant — no difference among the three groups (negative for clinical benefit)
Estimated mean % increase in ICH volume at 24 hSecondaryPlacebo: 26% · rFVIIa 20 μg/kg: 18% (P=0.09 vs placebo) · rFVIIa 80 μg/kg: 11% (P<0.001 vs placebo)
Absolute reduction in ICH growth (mL) vs placeboSecondaryrFVIIa 20 μg/kg: -2.6 mL (95% CI -5.5 to 0.3), P=0.08 · rFVIIa 80 μg/kg: -3.8 mL (95% CI -6.7 to -0.9), P=0.009
Total lesion (ICH+IVH+edema) growth at 24 hSecondary7 mL less with 80 μg/kg vs placebo (P=0.06); 72-h total lesion volumes similar across groups
Mortality at 3 monthsSecondaryPlacebo: ~19-20% · rFVIIa 20 μg/kg: ~20% · rFVIIa 80 μg/kg: ~20% · No mortality benefit
NIHSS at 90 daysSecondarySignificantly lower with 80 μg/kg vs placebo (small magnitude of difference)
Barthel Index at 90 daysSecondarySimilar median scores across three groups
mRS distribution at 90 daysSecondarySimilar shift distribution across groups (no benefit)
Post-hoc subgroup (age ≤70, ICH ≤60 mL, IVH ≤5 mL, treated ≤2.5 h; 19% of cohort)SecondaryAdjusted OR for poor 90-day outcome with rFVIIa 80 μg/kg = 0.28 (95% CI 0.08-1.06, P=0.03) — hypothesis-generating
Time-dependent hemostatic effect (post-hoc)SecondaryReduction in ICH growth larger with earlier treatment: -4.5 mL (≤3 h from onset) and -5.6 mL (≤2 h from onset) for 80 μg/kg vs placebo; no significant interaction test
Overall thromboembolic serious adverse eventsAdverseSimilar across three groups
Arterial thromboembolic SAEs (placebo vs 80 μg/kg)Adverse4% vs 9% (absolute +5%, P=0.04)
Venous thromboembolic eventsAdverseNo increase with rFVIIa vs placebo
Elevated cardiac troponin IAdversePlacebo 15% / 20 μg/kg 13% / 80 μg/kg 22%
ST-segment elevation myocardial infarctionAdversePlacebo 1.5% / 20 μg/kg 0.4% / 80 μg/kg 2.0%
CT evidence of acute cerebral infarctionAdversePlacebo 2.2% / 20 μg/kg 3.3% / 80 μg/kg 4.7%
Risk factors for thromboembolic SAE (post-hoc)AdverseAge (OR per 5-yr increase 1.1, 95% CI 1.0-1.2, P=0.02) and prior antiplatelet use (OR 1.9, 95% CI 1.1-3.0, P=0.01); rFVIIa treatment itself and prior thromboembolic disease were NOT independent risk factors

Subgroup Analysis

Prespecified subgroup analyses were not significant. Exploratory post-hoc analyses tested age (65/70/75/80 y cutoffs), baseline ICH volume (60 mL cutoff), IVH volume (5 mL cutoff), and time-to-treatment (2.0/2.5/3.0 h cutoffs). A composite subgroup — age ≤70, ICH ≤60 mL, IVH ≤5 mL, treated ≤2.5 h (19% of cohort) — showed adjusted OR 0.28 (95% CI 0.08-1.06, P=0.03) for poor 90-day outcome with 80 μg/kg rFVIIa; this hypothesis-generating signal motivated later ultra-early / spot-sign trials (FASTEST, SPOTLIGHT, STOP-IT, TRAIGE).


Criticisms

  • Baseline randomization imbalances disadvantaged the 80 μg/kg arm — higher rates of intraventricular hemorrhage (41% vs 29% placebo), coma (GCS 6-8), left ventricular hypertrophy, and larger total lesion volume — which likely biased the primary outcome against active treatment.
  • The placebo group had unexpectedly favorable outcomes (mortality 19%, poor outcome 24%) compared with the earlier phase 2b trial (mortality 29%, poor outcome 45%), reducing the absolute effect size rFVIIa could achieve.
  • Sponsor (Novo Nordisk) responsible for trial operations and initial data analysis; multiple author financial ties (consulting, lecture fees, employment, stock ownership).
  • Trial did not require CT-angiographic 'spot sign' or other markers of active bleeding to enrich for patients most likely to benefit from hemostatic therapy.
  • 4-hour treatment window may have been too long — only 17% treated within 2 hours; hemostatic effect and possible clinical benefit were largest with earliest treatment.
  • Broad inclusion allowed elderly patients (up to 97 years) at high risk of non-neurologic death, diluting any neurologic benefit.
  • Reliance on modified Rankin 5-6 as primary endpoint (rather than shift analysis across full mRS) may have missed smaller functional gains; NIHSS did show a small significant improvement with 80 μg/kg.
  • Post-hoc subgroup analyses (young, small ICH, minimal IVH, treated ≤2.5 h) are hypothesis-generating only and were not prespecified.

Funding

Novo Nordisk (Bagsvaerd, Denmark) — manufacturer of NovoSeven (rFVIIa); sponsor was responsible for trial operations and data analysis. Multiple authors reported employment by, consulting/lecture fees from, and/or stock holdings in Novo Nordisk.

Based on: FAST (New England Journal of Medicine, 2008)

Authors: Mayer SA, Brun NC, Begtrup K, ..., for the FAST Trial Investigators

Citation: N Engl J Med 2008;358(20):2127-2137. DOI: 10.1056/NEJMoa0707534

Content summarized and formatted by NeuroTrials.ai.