Background and Purpose: Fast and slow progressors of infarct growth due to a large vessel occlusion (LVO) are commonly observed when patients are considered for acute endovascular therapy. However, the incidence of these clinical phenotypes has not been formally quantified. We aimed to measure the frequency and distribution of fast versus slow infarct growth in LVO patients in different time strata after stroke onset.
Methods: Retrospective study of consecutive patients with acute LVO admitted to our comprehensive stroke center. Inclusion criteria were: 1. CTA or MRA evidence of intracranial ICA and/or MCA(M1) occlusion, with or without extracranial ICA occlusion. 2. Baseline CTP or MRI DWI before reperfusion therapy and within 25 hours of time of last seen well/stroke onset. Automated core volume measurements were performed using RAPID software (iSchemaView). Fast progressors were defined as having core volumes > 70 cc at early time quantiles (0 - 8 hours). Slow progressors were defined as having core volumes < 10 cc at later time quantiles (> 8 hours). Infarct growth rate was computed using ischemic core volume divided by time from stroke onset to core imaging. Descriptive statistics and univariate analysis were performed.
Results: A total of 115 patients were included with mean age 69+/-15.5 and NIHSS 16.5+/-7; 66% were females. Patients had MCA occlusion in 81% and intracranial ICA occlusion in 19% of cases. The median time from stroke onset to baseline core imaging was 9.1 hours (IQR, 4 - 14.7) and 53 patients were imaged before 8 hours. The median core volume was 10 (range 0 - 405 ml). In the early time epoch (< 8 hrs), 5/13 patients (38%) had core > 70ml by 3.65 hours (fast progressors). In the later time epoch (> 8 hrs), 29/62 patients (46%) had ischemic cores < 10ml at 25 hours from stroke onset. The median infarct growth rate was 2.6 ml/hr (0 - 14.6) in patients imaged before 8 hours and 1.2 ml/hr (0 - 5.48) in the later time epoch.
Conclusions: The incidence of fast and slow progressors of LVO stroke may be higher than previously anticipated depending on different time quantiles of stroke onset to imaging.