IV thrombolysis in wake-up stroke with DWI-FLAIR mismatch: evidence for extending the treatment window?

67F found unresponsive with left hemiplegia at 06:00, last known well at 22:00 (8-hour window). NIHSS 14. CT showed no hemorrhage. MRI (door-to-MRI 28 minutes):

DWI: acute right MCA territory infarct involving insula and frontal operculum (volume ~18 mL)
FLAIR: no corresponding hyperintensity (positive DWI-FLAIR mismatch)
MRA: right M1 occlusion with good collaterals

CTP:

Core: 15 mL
Penumbra: 85 mL (Tmax > 6s)
Mismatch ratio: 5.7

The WAKE-UP trial showed benefit of IV alteplase in DWI-FLAIR mismatch patients. But this patient also qualifies for mechanical thrombectomy based on DAWN/DEFUSE 3 imaging criteria.

Questions:

Do you give IV alteplase before thrombectomy in this scenario, or proceed directly to thrombectomy?
If giving alteplase, do you use the full 0.9 mg/kg dose or a reduced dose (as is common practice in Japan at 0.6 mg/kg)?
How do you interpret the absence of FLAIR changes at 8 hours: does this truly indicate < 4.5 hours from onset, or could it represent slow diffusion in well-collateralized territory?

Excellent case that highlights the evolving paradigm in stroke treatment. My approach at NIMHANS:

1. IV alteplase before thrombectomy: Yes, I give IV thrombolysis before thrombectomy in this scenario. The rationale:

WAKE-UP trial specifically validated DWI-FLAIR mismatch for IV thrombolysis decision-making
The "drip and ship" model showed no harm from bridging IV tPA in the SWIFT DIRECT and DIRECT-MT trials (though these tested known-onset strokes)
IV alteplase may lyse distal clot fragments and improve microvascular reperfusion even if the M1 occlusion requires mechanical intervention
Do not delay thrombectomy for alteplase completion. Start the infusion and proceed to angiography simultaneously.

2. Dosing: I use the standard 0.9 mg/kg (10% bolus, 90% infusion over 1 hour). I am aware of the J-MARS and ENCHANTED data supporting 0.6 mg/kg in Asian populations, but the ENCHANTED trial showed a trend toward inferior outcomes with low-dose tPA in large vessel occlusion specifically. For an M1 occlusion, I want maximum thrombolytic effect.

3. DWI-FLAIR mismatch interpretation: The absence of FLAIR signal at 8 hours from LKW is most likely because the true onset is more recent (probably 3-5 hours). FLAIR hyperintensity typically appears within 4-6 hours of stroke onset. However, you are correct that excellent collaterals can delay FLAIR changes. The CTP data here is reassuring: 15 mL core with 85 mL penumbra and mismatch ratio 5.7 confirms substantial salvageable tissue regardless of exact onset time.

Bottom line for this patient: IV alteplase 0.9 mg/kg + immediate thrombectomy. The imaging profile is ideal.

Thank you Dr. Nair. Here is how we handled this case at University of Tokyo Hospital:

We administered alteplase 0.6 mg/kg (the standard Japanese dose per J-MARS protocol) and proceeded immediately to mechanical thrombectomy. The rationale for 0.6 mg/kg in our practice:

ENCHANTED showed non-inferiority for 0.6 vs 0.9 mg/kg overall
In Japan, we have strong ICH safety data at 0.6 mg/kg with lower symptomatic ICH rates (2.1% vs 4.5% in NINDS)
Since this patient is going straight to thrombectomy anyway, the incremental benefit of full-dose systemic lysis is marginal

Thrombectomy result: TICI 2c reperfusion achieved with single-pass Solitaire retrieval. Door to groin: 62 minutes. Door to reperfusion: 78 minutes.

24-hour NIHSS: 3 (from 14 at presentation) 7-day MRI: Final infarct volume 22 mL (essentially limited to the initial DWI core)

On the DWI-FLAIR mismatch question: We published a retrospective series (n=84) last year showing that DWI-FLAIR mismatch has a positive predictive value of 83% for symptom onset within 4.5 hours. The remaining 17% were patients with excellent collateral grades (ASITN/SIR 3-4), which delays tissue injury and FLAIR signal development. In this patient with good collaterals on MRA, the mismatch is likely genuine, but the collaterals are contributing to the delayed FLAIR appearance.

2 Answers