Background:
Early reperfusion is associated with improved clinical outcome in acute ischemic stroke; however, there is no consensus regarding which perfusion parameter may best serve as a marker of clinical improvement. We compared three commonly used MRI perfusion parameters, mean transit time (MTT), time-to-peak (TTP), and Tmax, to identify which method of measuring reperfusion best predicted clinical improvement.
Methods:
Acute ischemic stroke patients underwent two MR scans: within 4.5 hours (tp1) and at 6 hours (tp2) after stroke onset. Co-registered MTT, TTP, and Tmax maps were generated to measure regions of perfusion deficit at tp1 and tp2. Perfusion deficit was defined as prolongation of MTT, TTP, or Tmax beyond four pre-specified thresholds for each parameter (4 thresholds were chosen to ensure results were not spuriously based on one threshold). Commonly-used thresholds (relative to contralateral median) were selected for each parameter: for MTT: >3, 4, 5, or 6 seconds (s), for TTP: >2, 4, 6, or 8s, and for Tmax: >2, 4, 6, and 8s. The volume of reperfusion (Vreperf) was defined as the volume of tissue with perfusion deficit at tp1 and no perfusion deficit at tp2. Clinical improvement was defined as: Admission NIH Stroke Scale (NIHSS) - 1 month NIHSS (ΔNIHSS). A multivariable linear regression model identified if Vreperf as measured by MTT, TTP, or Tmax was an independent predictor of clinical improvement after adjusting for patient age, admission NIHSS, tPA treatment, and volume of tp1 perfusion deficit.
Results:
Thirty-nine acute ischemic stroke patients were prospectively scanned at 2.8±.8hr (tp1) and 6.4±.4hr (tp2) after stroke onset (mean age=64, 44% female, 36% Black, mean NIHSS=14, 74% received IV tPA). Across the four thresholds, mean volume of perfusion deficit ranged from 58-96ml for MTT, 56-116ml for TTP, and 51-113ml for Tmax. Mean Vreperf ranged from 15-22ml for MTT, 15-23ml for TTP, and 14-21ml for Tmax. In the multivariable linear regression analysis, after adjusting for age, admission NIHSS, tPA treatment, and volume of tp1 perfusion deficit, Vreperf predicted ΔNIHSS for MTT=4s (p=0.007), MTT=5s (p=0.005), and MTT=6s (p=0.010), whereas Vreperf did not predict ΔNIHSS for any TTP or Tmax threshold (
Table
).
Conclusion:
Reperfusion, defined by MTT, was an independent predictor of clinical improvement, while reperfusion defined by TTP and Tmax were not. Therefore, MTT may be the best time-based perfusion parameter to define clinically-relevant reperfusion after stroke.
Perfusion Parameter Thresholds That Discriminate Ischemic Core Vary with Time from Onset in Acute Ischemic Stroke