Infarct volume versus NIHSS score for ischemic and bleeding risk stratification after acute ischemic stroke.

BACKGROUND: Whether infarct volume refines assessment of recurrent ischemic stroke (IS) and intracerebral hemorrhage (ICH) risk following acute IS beyond the National Institutes of Health Stroke Scale (NIHSS) remains uncertain. We compared their performance in ischemic-hemorrhagic risk stratification. METHODS: We included 10,545 patients with acute IS from the Third China National Stroke Registry (CNSR-III) with diffusion-weighted imaging/apparent diffusion coefficient-derived infarct volume and NIHSS data. Outcomes were 90-day recurrent IS and ICH, analyzed separately. Two reciprocal Cox models were used: one with infarct volume adjusted for NIHSS, and another with NIHSS adjusted for infarct volume. Stratified analyses, restricted cubic splines, and grouped threshold analyses were exploratory. In the GOLDEN BRIDGE II study, recurrent IS and all bleeding events were assessed for external consistency. RESULTS: In CNSR-III, larger infarct volume was independently associated with recurrent IS [hazard ratio (HR)=1.35, 95% confidence interval (CI) 1.25-1.47] and ICH (HR=2.48, 95% CI 2.11-2.91) after NIHSS adjustment, whereas NIHSS lost significance after infarct-volume adjustment (IS: HR=1.05, 95% CI 0.89-1.24; ICH: HR=1.11, 95% CI 0.78-1.58). ICH risk rose around 9.0 ml, with increased hazard at 9.0-9.9 ml (HR=3.61, 95% CI 1.29-10.07), while IS risk increased without a sharp rise (HR=1.15, 95% CI 0.98-1.35). Patients with NIHSS score ≥4 and infarct volume <9.0 ml had IS risk comparable to NIHSS score ≤3 but potentially lower ICH risk (HR=0.63, 95% CI 0.38-1.02). Infarct volume 9.0-17.9 ml increased ICH risk (HR=1.96, 95% CI 1.06-3.63). Infarct volume was <9.0 ml in 66.7% of NIHSS score 6-15 and 25.4% of NIHSS score ≥16. GOLDEN BRIDGE II confirmed principal findings, with infarct volume associated with recurrent IS (HR=1.37, 95% CI 1.25-1.49) and bleeding events (HR=1.53, 95% CI 1.31-1.78). CONCLUSIONS: Infarct volume better characterizes post-stroke ischemic-hemorrhagic risk than NIHSS. Prospective validation is required before using infarct volume to guide antiplatelet selection.