SITS-MOST
(2007)Objective
To confirm the safety and efficacy of intravenous alteplase within 3 h of acute ischaemic stroke onset in routine European clinical practice, including centres with little prior thrombolysis experience.
Study Summary
• sICH per Cochrane/NINDS definition at 7 d 7.3% (468/6438; 6.7–7.9) vs 8.6% (40/465; 6.3–11.6) in pooled RCTs; sICH per ECASS definition 4.6% (296/6442) vs 8.8% in ECASS II.
• 3-month mortality 11.3% (701/6218; 10.5–12.1) vs 17.3% (83/479; 14.1–21.1) in pooled RCTs.
• Complete recovery (mRS 0–1) at 3 months 38.9% (2386/6136; 37.7–40.1) vs 42.3% in RCTs; functional independence (mRS 0–2) improved ~5% over RCTs.
• Safety and effectiveness were preserved in new/inexperienced centres: sICH (ECASS) 4.6% and mRS 0–1 40.7% in new centres vs 4.6% and 38.3% in experienced centres.
Intervention
Intravenous alteplase 0.9 mg/kg within 3 h of ischaemic stroke onset (per European summary of product characteristics).
Inclusion Criteria
Age 18–80 y; ischaemic stroke with treatment initiated ≤3 h from onset; NIHSS <25; no contraindications per European alteplase SmPC.
Study Design
Arms: Single-arm prospective observational cohort of alteplase-treated patients; safety/efficacy benchmarked against pooled ATLANTIS/ECASS/NINDS RCT data (placebo and alteplase arms within 3 h).
Patients per Arm: 6483 alteplase-treated patients (SITS-MOST); pooled RCT comparator n=464 alteplase / n=465 placebo (0–3 h stratum).
Outcome
• Primary efficacy: 3-month mortality 11.3% vs 17.3% in pooled RCTs.
• Secondary: mRS 0–2 at 3 months ~55% (about 5 percentage points higher than pooled RCTs); mRS 0–1 38.9% vs 42.3%.
• Outcomes similar between experienced and new centres.
Clinical Question
In routine European clinical practice, is intravenous alteplase given within 3 h of acute ischaemic stroke onset as safe and effective as in the randomised controlled trials, even at centres with limited previous thrombolysis experience?
Bottom Line
Post-marketing observational data from 6483 patients across 285 European centres show that IV alteplase within 3 h of ischaemic stroke onset has a safety profile at least as favourable as the pooled RCTs (sICH per Cochrane 7.3% vs 8.6%; 3-month mortality 11.3% vs 17.3%) and remains safe and effective in centres new to stroke thrombolysis.
Major Points
- Largest post-marketing safety study of IV alteplase in acute ischaemic stroke (n=6483 across 14 EU/associated countries, 2002–2006).
- Symptomatic ICH per SITS-MOST protocol (parenchymal haemorrhage type 2 with NIHSS deterioration ≥4) at 24 h: 1.7% (95% CI 1.4–2.0); fatal sICH 0.28%.
- sICH per Cochrane/NINDS definition at 7 d 7.3% (95% CI 6.7–7.9) vs 8.6% (6.3–11.6) in pooled RCTs; sICH per ECASS definition 4.6% vs 8.8% in ECASS II.
- 3-month mortality lower than in pooled RCTs (11.3% vs 17.3%) despite comparable baseline characteristics.
- Complete recovery (mRS 0–1) at 3 months 38.9% (SITS-MOST) vs 42.3% (pooled RCTs); functional independence (mRS 0–2) ~5 percentage points higher in SITS-MOST.
- Approximately half of participating centres were new to stroke thrombolysis; sICH rates (4.6% ECASS) and mRS 0–1 (40.7% vs 38.3%) were similar between new and experienced centres, supporting safe implementation in less experienced units.
- Basis for EU regulatory reassessment of alteplase for acute ischaemic stroke; catalyst for the ECASS III extension of the treatment window.
Study Design
- Study Type
- Prospective, open-label, multicentre, multinational observational monitoring study (single-arm cohort embedded within SITS-ISTR register)
- Randomization
- No
- Blinding
- Open-label (no blinding – observational registry)
- Sample Size
- 6483
- Follow-up
- 3 months
- Centers
- 285
- Countries
- 14 European countries (EU member states as of 2002 plus Norway and Iceland; ~50% of centres new to stroke thrombolysis)
Primary Outcome
Definition: Symptomatic intracerebral haemorrhage (SITS-MOST/PH2 definition: local or remote parenchymal haemorrhage type 2 on 22–36 h imaging with NIHSS deterioration ≥4 or death) within 24 h; and death within 3 months
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| - | - | - | - |
Limitations & Criticisms
- Observational, single-arm design with no concurrent control group – comparisons rely on historical pooled RCT data (ATLANTIS/ECASS/NINDS), which limits causal inference.
- Enrollment restricted to patients ≤80 years, NIHSS <25, and ≤3 h from onset per EU label – findings do not address older, more severe, or later-window patients.
- Potential selection and reporting bias despite source-data verification of ≥10% of records; participating centres self-selected and were required to meet stroke-unit standards.
- Follow-up data on modified Rankin Score at 3 months missing in ~5% of patients.
- Funded by Boehringer Ingelheim (manufacturer of alteplase); most senior authors received compensation from the sponsor.
- sICH rate differences across definitions (1.7% SITS-MOST/PH2 vs 4.6% ECASS vs 7.3% Cochrane/NINDS) illustrate how definition choice heavily influences the reported safety signal.
- Analyses were not powered for subgroup or covariate-adjusted comparisons; a multivariate analysis of centre experience was noted as 'underway' but not reported in this paper.
Citation
Lancet 2007;369(9558):275–282. DOI: 10.1016/S0140-6736(07)60149-4