BEYOND TICI 3: POST-SUCCESSFUL THROMBECTOMY DWI EXPANSION PREDICTS POOR STROKE RECOVERY

초록

Background and Aims: Complete reperfusion (TICI 3) after endovascular therapy (EVT) for large-vessel occlusion (LVO) stroke does not always halt ischemic damage. We quantified post-recanalisation infarct growth on diffusion-weighted MRI (DWI) and explored its determinants and clinical impact. Methods: We retrospectively studied 189 consecutive LVO patients with baseline CT perfusion, “early” DWI ≤24 h (median 1.7 h) and “delayed” DWI 24–168 h (median 92.8 h) after EVT. Infarct volumes were extracted by automated JLK-CTP/DWI software. Growth was the absolute volume change between scans. Predictors were assessed with robust linear regression; associations with 90-day modified Rankin Scale (mRS) were tested using multivariable logistic regression and mediation analysis. Results: Median infarct volume increased from 4.8 mL to 11.7 mL, yielding a median growth of 4.4 mL; 26 patients (13.8 %) showed lesion shrinkage. Larger baseline ischemic core (rCBF < 30 %) (β = 0.187 ± 0.025 mL per mL, p < 0.001) and higher admission glucose (β = 0.049 ± 0.018 mL per mg/dL, p = 0.007) independently predicted greater growth. Each decile increase in infarct growth lowered the odds of functional independence (mRS 0–2) at 90 days (adjusted OR 0.81, 95 % CI 0.68–0.95) and excellent outcome (mRS 0–1) (OR 0.80, 95 % CI 0.69–0.93). Infarct growth partially mediated the association between baseline core volume and 90-day mRS (p = 0.02). Models incorporating delayed DWI volume discriminated outcomes better than those using early DWI or CTP core alone (Figure). Conclusions: Substantial infarct enlargement can occur hours to days after technically successful EVT, driven chiefly by larger baseline core and hyperglycaemia and translating into worse functional recovery. These findings highlight the need for adjunct neuroprotective strategies and support delayed DWI lesion volume as a robust surrogate end-point for future acute-stroke trials.