Abstract Background and aims The location-specific optimal cut point of ultra-early hematoma growth speed (UHG) for prognosticating acute intracerebral hemorrhage (ICH) is unknown. Methods Secondary analyses of participant data are from the INTERACT (Intensive Blood Pressure Reduction in Acute Intracerebral Hemorrhage Trials) 1-3 studies. UHG was defined as hematoma volume (mL) divided by onset-to-CT time (hours). Cutoffs for UHG in whole and location-specific ICH with optimal sensitivity and specificity for 90-day or 180-day clinical outcomes (death or major disability defined as modified Rankin Scale scores 3-6, and death alone) were identified using receiver operating characteristic curves. Results There were 4109 ICH patients (age, 62.8 ± 12.6 years; female, 36.0%; UHG, 5.55 0.69-29.65 mL/h; location lobar 10.9%, deep 82.2%, infratentorial 6.9%) included in the main analyses. Receiver operating characteristic curve analysis identified an optimal UHG cutoff of 6.04 mL/h for death or major disability (area under the curve AUC=0.81) and 8.74 mL/h for death alone (AUC=0.79). Location-specific optimal UHG cutoffs for death or major disability were 8.73 mL/h (AUC=0.82) for lobar, 5.77 mL/h (AUC=0.79) for deep, and 4.35 mL/h (AUC=0.81) for infratentorial ICH, whereas corresponding cutoffs for death alone were 7.07 mL/h (AUC=0.74), 12.77 mL/h (AUC=0.81), and 7.91 mL/h (AUC=0.83), respectively. Conclusions In patients with ICH, baseline UHG cutoffs of 6.04 mL/h and 8.74 mL/h were optimal for predicting death or major disability and death alone, respectively, at 90 or 180 days, with values varying by hematoma location. Conflict of interest Dr You holds the National Natural Science Foundation of China (82471226), Discipline Construction Program of the Second Affiliated Hospital of Soochow University (XKTJ-XK202401 and XKTJ-RC202412). Dr Anderson holds a the National Health and Medical Research Council (NHMRC) of Australia Senior Investigator Fellowship and reports grant funding from NHMRC, Penumbra, and Takeda China. The other authors have nothing to disclose.
You et al. (Fri,) studied this question.