← Back
NeuroTrials.ai
Neurology Clinical Trial Database

THAWS

Thrombolysis for Acute Wake-Up and Unclear-Onset Strokes With Alteplase at 0.6 mg/kg: A Randomized Controlled Trial

Share Share Copied! Compare

Year of Publication: 2020

Authors: Koga M, Yamamoto H, Inoue M, ..., Toyoda K; for the THAWS Trial Investigators

Journal: Stroke

Citation: Koga M, et al. Stroke. 2020;51:1530–1538. DOI: 10.1161/STROKEAHA.119.028127.

Link: https://www.ahajournals.org/doi/10.1161/STROKEAHA.119.028127

PDF: https://www.ahajournals.org/doi/reader/1...EAHA.119.028127

Clinical Question

In patients with acute ischemic stroke of unknown onset time and DWI-FLAIR mismatch on MRI, does intravenous alteplase at the Japan-approved dose of 0.6 mg/kg improve 90-day functional outcome compared with standard medical treatment?

Bottom Line

THAWS was terminated early (131/300 patients) after the WAKE-UP trial demonstrated efficacy of full-dose alteplase in the same population; no significant difference in favorable outcome (mRS 0–1 at 90 days) was found between alteplase (47.1%) and control (48.3%; RR 0.97, 95% CI 0.68–1.41, P=0.892). Symptomatic intracranial hemorrhage was rare (1.4% vs 0%; P>0.999) and mortality was similar (2.8% vs 3.3%; P>0.999), confirming the safety of 0.6 mg/kg alteplase in this setting, though early termination and underpowering preclude definitive efficacy conclusions.

Major Points

  • THAWS was an investigator-initiated, multicenter, randomized, open-label, blinded-endpoint trial conducted at 40 centers across Japan from May 2014 to July 2018, testing alteplase at the Japan-approved dose of 0.6 mg/kg (rather than the global standard 0.9 mg/kg) in DWI-FLAIR mismatch stroke with unknown onset.
  • The trial was stopped prematurely on July 10, 2018 after the WAKE-UP trial published positive results for 0.9 mg/kg alteplase in the same population; only 131 of 300 planned patients (43.7%) were enrolled.
  • The primary endpoint—mRS 0–1 at 90 days—was not significantly different: 47.1% (32/68) in the alteplase group vs 48.3% (28/58) in the control group (RR 0.97, 95% CI 0.68–1.41, P=0.892).
  • Symptomatic intracranial hemorrhage (defined as NIHSS increase ≥4 plus PH-2 on MRI at 22–36 hours) occurred in only 1 of 71 patients in the alteplase group (1.4%) and 0 of 60 in the control group (RR infinity, 95% CI 0.06 to infinity, P>0.999), confirming safety of the 0.6 mg/kg dose.
  • Mortality at 90 days was 2.8% (2/71) in the alteplase group vs 3.3% (2/60) in controls (RR 0.85, 95% CI 0.06–12.58, P>0.999); no major extracranial bleeding occurred in either group.
  • The only significant imaging finding was earlier recanalization: 73.7% (14/19) in the alteplase group vs 40.9% (9/22) in the control group (RR 1.80, 95% CI 1.02–3.64, P=0.04), though the number of patients with vascular imaging available was small (n=41).
  • The control group performed better than expected (mRS 0–1: 48.3% vs 41.8% in WAKE-UP controls), likely due to the open-label design allowing early aggressive antithrombotic therapy: 85% of controls received any antithrombotic within 24 hours vs only 13% in the alteplase group, including 23% vs 3% dual antiplatelet therapy.
  • A significant prespecified subgroup interaction was found with premorbid antithrombotic medication (P<0.01): in patients already on antithrombotics (n=40), alteplase was associated with more favorable outcomes, though this subgroup was small.
  • The trial enrolled many mild-to-moderate strokes (median NIHSS 7 both groups) and few large vessel occlusions (27–36% had any vessel occlusion on MRA), partly because thrombectomy became standard during the enrollment period and those patients were diverted, reducing the alteplase treatment effect.
  • The 0.6 mg/kg dose used in THAWS (Japan standard since J-ACT trial) is 33% lower than the 0.9 mg/kg dose tested in WAKE-UP and MR-WITNESS; nevertheless, sICH rates in the alteplase group were comparable (1.4% in THAWS, 1.3% in MR-WITNESS, 2.0% in WAKE-UP) and 90-day mortality was lowest in THAWS (2.8% vs 8.8% MR-WITNESS vs 4.1% WAKE-UP).

Design

Study Type: Investigator-initiated, multicenter, randomized, open-label, blinded-endpoint (PROBE design), parallel-group controlled trial

Randomization: 1:1 minimization scheme stratified by NIHSS severity (≤11 vs >11); allocated via in-house interactive email-based system at time of enrollment

Blinding: Open-label treatment allocation (both patients and investigators unblinded); primary and all efficacy/safety outcomes assessed by independent certified neurologists, neurosurgeons, nurses, or clinical research coordinators blinded to treatment allocation

Enrollment Period: May 1, 2014 – July 10, 2018

Follow-up Duration: 90 days

Centers: 40

Countries: Japan

Sample Size:

  • Planned: 300
  • Enrolled: 131
  • Alteplase Group (randomized): 70
  • Control Group (randomized): 61
  • Safety Analysis Set - Alteplase: 71
  • Safety Analysis Set - Control: 60
  • ITT - Alteplase: 70
  • ITT - Control: 61
  • Per-Protocol - Alteplase: 67
  • Per-Protocol - Control: 57
  • As-Treated - Alteplase: 71
  • As-Treated - Control: 60

Power Calculation: 278 patients required for 90% power (1-sided alpha=2.5%) to demonstrate relative effect of alteplase is >0.5 of combined relative effect from historical thrombolysis studies, assuming 30% favorable rate with alteplase and 20% with control

Analysis: Primary: intention-to-treat. Secondary: per-protocol (excluding major protocol violations) and as-treated. Multiple imputation and marginal model with linearization-based method used as sensitivity analyses. SAS/STAT v9.4. P<0.05 (2-sided) considered significant.

Inclusion Criteria

  • Acute ischemic stroke with symptoms on awakening or with unknown time of onset
  • Last-known-well to presentation interval >4.5 hours
  • Presentation within 4.5 hours after symptom recognition/discovery
  • Age >=20 years
  • NIHSS score >=2 (original protocol: >=5; revised to >=2 in August 2015) and <=25
  • Premorbid mRS score within eligibility range per Japanese alteplase guidelines
  • MRI showing acute ischemic lesion on DWI with no marked corresponding hyperintensity on FLAIR (DWI-FLAIR mismatch / negative FLAIR pattern), using standard FLAIR settings per WAKE-UP imaging guidelines
  • Met all standard eligibility criteria for intravenous alteplase in Japan per Japan Stroke Society 2012 guidelines EXCEPT the time last-known-well criterion
  • Time limit: Initially <=12 hours since last-known-well (May 2014 protocol); revised to no upper time limit in August 2015 to match WAKE-UP trial criteria
  • Note: Patients with confirmed acute ischemic stroke and negative FLAIR who did NOT display an abnormal DWI signal were also eligible

Exclusion Criteria

  • NIHSS <2 (original cutoff <5, revised to <2 in August 2015) or NIHSS >25
  • Any contraindication to MRI (e.g., cardiac pacemaker)
  • Planned or anticipated treatment with surgery or endovascular reperfusion strategies (thrombectomy)
  • Pregnancy, lactating, or potentially pregnant
  • Life expectancy <=6 months per investigator judgment
  • Intracranial hemorrhage on baseline MRI
  • Large infarct: DWI-ASPECTS <=4 in the MCA territory
  • Visual lesion volume >50% of the ACA territory
  • Visual lesion volume >50% of the PCA territory
  • Infarct involving >50% of the brainstem
  • Infarct involving >50% of a unilateral cerebellar hemisphere
  • Any standard contraindication to intravenous alteplase per Japanese guidelines

Baseline Characteristics

CharacteristicAlteplase Group (n=70)Control Group (n=61)
Age (mean +/- SD), years73.2 +/- 12.475.8 +/- 11.9
Female sex, n (%)25 (36%)30 (49%)
Hypertension, n (%)49 (70%)41 (67%)
Diabetes mellitus, n (%)14 (20%)12 (20%)
Dyslipidemia, n (%)23 (33%)23 (38%)
Atrial fibrillation, n (%)27 (39%)21 (34%)
History of ischemic stroke or TIA, n (%)8 (11%)14 (23%)
Wake-up stroke (sleep), n (%)53 (76%)40 (66%)
Unknown onset - other reason, n (%)17 (24%)21 (34%)
NIHSS score, median (IQR)7 (4-13)7 (5-12)
Any vessel occlusion on MRA, n (%)19 (27%)22 (36%)
ICA occlusion, n (%)1 (1%)2 (3%)
MCA M1 occlusion, n (%)6 (9%)8 (13%)
MCA M2 occlusion, n (%)11 (16%)11 (18%)
PCA occlusion, n (%)1 (1%)0
Basilar artery occlusion, n (%)01 (2%)
DWI-ASPECTS, median (IQR)9 (8-10)9 (8-10)
DWI lesion volume (mL), median (IQR)3.5 (0.8-14.2)1.9 (0.3-11.1)
Time last-known-well to symptom recognition, median (IQR), hours7.0 (5.5-9.0)7.0 (5.5-9.0)
Time symptom recognition to MRI, median (IQR), hours2.0 (1.5-2.8)2.0 (1.5-2.8)
Time MRI start to randomization, median (IQR), minutes52.0 (35.7-70.3)44.1 (33.0-62.4)
Time symptom recognition to randomization, median (IQR), hours3.1 (2.4-3.7)2.9 (2.3-3.8)
Time last-known-well to randomization, median (IQR), hours10.2 (8.2-12.2)10.3 (7.7-11.8)

Arms

FieldAlteplase 0.6 mg/kgControl
InterventionIntravenous alteplase at 0.6 mg/kg total dose: 10% given as IV bolus, remaining 90% by 60-minute continuous infusion. Treatment initiated within 4.5 hours of symptom recognition/waking, and within 60 minutes of MRI completion. Antithrombotic agents prohibited for first 25 hours post-treatment. Thrombolytics (urokinase, monteplase, tenecteplase) prohibited for 90-day study period.Antithrombotic therapy per attending physician discretion using 1-3 agents: oral aspirin 160-300 mg/day, oral clopidogrel 75 mg/day, intravenous argatroban, or intravenous unfractionated heparin (combination of argatroban and heparin excluded). In practice: any antithrombotic in 85%, oral antiplatelet in 49%, IV anticoagulant in 66% of control patients within 24 hours. Thrombolytics prohibited for 90-day study period.
n7061

Outcomes

OutcomeTypeControlInterventionHR / OR / RRP-value
Favorable outcome: mRS score 0-1 at 90 days after stroke onsetPrimary28/58 (48.3%; 95% CI 35.0-61.8%)P=0.892
Median mRS score at 90 daysSecondary2 (IQR 0-3)P=0.67
mRS 0-2 at 90 daysSecondary35/58 (60.3%; 95% CI 46.6-73.0%)P=0.862
Categorical shift in NIHSS score at 24 hours from baselineSecondary-2.8 +/- 3.9 (n=61)P=0.85
Categorical shift in NIHSS score at 7 days from baselineSecondary-4.3 +/- 4.2 (n=60)P=0.44
Symptomatic intracranial hemorrhage at 22-36 hours (SITS-MOST: PH-2 plus NIHSS increase >=4 from baseline)Safety0/60 (0%)P>0.999
Major extracranial bleedingSafety0/60 (0%)Not applicable
Death at 90 daysSafety2/60 (3.3%) [causes: heart failure n=1, gastric cancer n=1]P>0.999
Serious adverse eventsSafety6/60 (10.0%)P=0.632

Criticisms

  • Premature termination at 43.7% of planned enrollment (131/300) severely underpowered the trial; early stopping precludes definitive efficacy conclusions
  • Open-label treatment allocation (no placebo) introduced treatment bias: early antithrombotic initiation in 85% of controls within 24 hours likely improved control group outcomes, diminishing apparent alteplase treatment effect
  • Lower dose of alteplase (0.6 mg/kg, Japan standard) than the 0.9 mg/kg used in WAKE-UP and MR-WITNESS makes direct dose comparison impossible and limits generalizability outside Japan
  • Exclusion of thrombectomy-eligible patients (45 screened patients excluded) introduced selection bias toward milder strokes with predominant small-vessel disease, for which thrombolysis has limited benefit
  • Only 27-36% of enrolled patients had any vessel occlusion on MRA; thrombolysis is most beneficial in large vessel occlusion, so the enrolled population may not be the optimal target
  • Potential stroke mimic inclusion: 39 patients had DWI-ASPECTS of 10 (normal DWI), with 9 showing no FLAIR lesion at 7 days, suggesting some cases may represent stroke mimics or TIA
  • Imaging follow-up for recanalization was only available in 41 patients (31% of total), limiting validity of the significant recanalization finding
  • Stratification by NIHSS <=11/>11 differed from WAKE-UP (<=10/>10) and did not include age-based stratification used in WAKE-UP

Funding

Japan Agency for Medical Research and Development (AMED; grant numbers 19ek0210091h0003 and 19lk0201094h0001); Ministry of Health, Labour, and Welfare of Japan; Mihara Cerebrovascular Disorder Research Promotion Fund. No pharmaceutical industry funding reported.

Based on: THAWS (Stroke, 2020)

Authors: Koga M, Yamamoto H, Inoue M, ..., Toyoda K; for the THAWS Trial Investigators

Citation: Koga M, et al. Stroke. 2020;51:1530–1538. DOI: 10.1161/STROKEAHA.119.028127.

Content summarized and formatted by NeuroTrials.ai.