Pushing the Clock: The Expanding Time Window for Ischemic Stroke

In the mid-1990s, stroke treatment entered a new era with the approval of intravenous tPA within 3 hours of onset. Over the next three decades, progressive advances in imaging, patient selection, and trial design stretched the therapeutic window far beyond what once seemed possible. Here’s how we got from 3 hours to >24 hours—and what it means for clinical practice today.

3 Hours: The NINDS tPA Trial (1995)

The original NINDS trial established that IV alteplase given within 3 hours of stroke onset improved the odds of a favorable outcome (mRS 0–1 at 90 days) by 12% absolute (NNT ≈ 8) despite a 6.4% risk of symptomatic ICH(1).

4.5 Hours: ECASS III (2008)

ECASS III extended the window to 4.5 hours. Patients saw an 8% absolute improvement in functional outcome (mRS 0–1), with ICH risk rising to 7.9%(2). Originally excluded groups (age >80, NIHSS >25, prior stroke + diabetes) are now variably included.

9 Hours: EXTEND (2019)

EXTEND used perfusion imaging to identify mismatch up to 9h from onset. Alteplase improved mRS 0–1 rates by ~6% (35.4% vs. 29.5%)(3), with ICH risk ~6%.

Clinical Pearl: CT perfusion or MRI mismatch is now key for late thrombolysis. Widely adopted in comprehensive centers.

Unknown Onset: WAKE-UP (2018)

WAKE-UP used DWI/FLAIR mismatch on MRI to estimate onset within 4.5h in patients with unknown onset time. It improved outcomes: mRS 0–1 in 53% vs. 42% (NNT ≈ 9)(4), with sICH 2%.

24 Hour: Posterior Circulation — EXPECTS

EXPECTS evaluated alteplase in posterior circulation strokes 4.5–24h from onset (NIHSS ≥1, PC-ASPECTS ≥7), excluding planned thrombectomy. Alteplase improved mRS 0–2 (89.6% vs. 72.6%; RR 1.16; P = 0.01) with low sICH (1.7%).

24 Hour: With LVO – TIMELESS, CHABLIS-T II & TRACE III

  • TIMELESS: Tenecteplase 4.5–24h, with or without EVT. Neutral overall, but benefit in proximal LVO with good collaterals(10).
  • CHABLIS-T II: Alteplase in anterior LVO with CTP mismatch (mRS 0–2: 47% vs. 20%)(8).
  • TRACE III: Alteplase in anterior LVO not eligible for EVT, selected by CTP (mRS 0–2: 52.4% vs. 33.3%; p=0.01)(9).

Clinical Pearl: All three trials support tissue-based thrombolysis up to 24h in select LVO patients, especially when perfusion mismatch is present.

24 Hour: Without LVO – HOPE Trial

HOPE studied tenecteplase 4.5–24h in patients without LVO but with perfusion mismatch. mRS 0–1 at 90 days occurred in 39.7% vs. 29.0% with standard care (adjusted RR 1.37; P=0.047). sICH was low (1.3% vs. 0.7%).

Clinical Pearl: HOPE is the first RCT showing late-window thrombolysis benefits in non-LVO strokes using tenecteplase and perfusion imaging.

Conclusion: From Clock to Core

Physiologic imaging has extended thrombolysis from a fixed clock-based approach to patient-specific selection—enabling safe and effective treatment well beyond traditional windows.

🔹 Bottom Line: Extended Window

  • Unknown onset: Use MRI DWI-FLAIR mismatch (WAKE-UP strategy) to identify strokes < 4.5h.
  • 9-hour onset: CT perfusion–based mismatch (EXTEND) allows alteplase use in selected patients.
  • 24-hour window: Perfusion-guided thrombolysis is emerging for both LVO and non-LVO, expected to be endorsed by future guidelines.
Trial Time Window LVO Intervention Imaging Outcome sICH
NINDS 0–3h Not specified IV tPA NCCT mRS 0–1: 39% vs 26% 6.4%
ECASS III 3–4.5h Not specified IV tPA NCCT mRS 0–1: 52.4% vs 45.2% 7.9%
EXTEND 4.5–9h Mixed IV tPA CTP or MRI mRS 0–1: 35.4% vs 29.5% 6.2%
TIMELESS 4.5–24h LVO IV TNK CTP Neutral; subgroup benefit 3.2%
CHABLIS-T II 6–24h Anterior LVO IV tPA CTP mRS 0–2: 47% vs 20% Not stated

References

  1. NINDS tPA Stroke Study Group. N Engl J Med. 1995;333(24):1581–1587.
  2. Hacke W, et al. N Engl J Med. 2008;359:1317–1329.
  3. Ma H, et al. N Engl J Med. 2019;380:1795–1803.
  4. Thomalla G, et al. N Engl J Med. 2018;379:611–622.
  5. Xu G, et al. CHABLIS-T II. JAMA. 2024.
  6. Zhao X, et al. TRACE III. Lancet Neurol. 2024.
  7. TIMELESS Investigators. JAMA. 2024.
  8. HOPE Investigators. Stroke. 2025.