No Reflow at Perfusion Imaging: Thresholds That Distinguish Functional and Tissue Outcome Groups in Acute Stroke

Background Poor outcome after successful endovascular thrombectomy (EVT) may reflect microvascular dysfunction, or no-reflow phenomenon, but operational definitions are lacking. Purpose To identify the most informative candidate no-reflow definition at 24-hour post-EVT perfusion imaging for outcome assessment. Materials and Methods This secondary analysis used data from two prospective multicenter imaging registries of patients with acute stroke who were recruited between June 2016 and July 2022, achieved modified thrombolysis in cerebral infarction grade 2c or 3 recanalization after EVT, and underwent pre-EVT and 24-hour post-EVT CT or MR perfusion imaging. Candidate no-reflow definitions were evaluated across eight perfusion maps and their corresponding stepwise thresholds. Delay time (DT) quantifies the delay in contrast material arrival between arterial input and brain tissue. For primary screening, no-reflow status was defined as any residual hypoperfusion lesion for each map-threshold combination. The best-performing combination was evaluated using residual volume and reperfusion index cutoffs. The primary outcome was poor 90-day functional outcome (modified Rankin Scale score >2); infarct growth was secondary. Associations with the outcome were evaluated using multivariable Firth penalized logistic regression. Results The primary cohort included 67 patients (median age, 69 years [IQR, 56-77 years]; 41 male), and the independent cohort comprised 54 patients (median age, 69 years [IQR, 62-78 years]; 28 female). In the primary cohort, lesions with DT longer than 4 seconds showed the strongest discrimination for poor 90-day functional outcome (area under the receiver operating characteristic curve, 0.72 [95% CI: 0.57, 0.86]; P < .001). According to this definition, 14 of 67 patients (21%) had no reflow. Poor functional outcome occurred in 11 of 14 patients (79%) with no reflow and 11 of 53 (21%) with reflow (odds ratio, 12.14 [95% CI: 3.38, 54.63]; P < .001). Infarct growth was greater with no-reflow status than with reflow (median, 24.1 mL [IQR, 3.7-40.7 mL] vs 0 mL [IQR, -10.1 to 6.1 mL]; P < .001). Findings were consistent in the independent cohort. Conclusion The presence of a lesion with DT longer than 4 seconds at 24-hour post-EVT perfusion imaging appeared to be the most informative candidate no-reflow definition. © RSNA, 2026 Supplemental material is available for this article. See also the editorial by Gemmete in this issue.