Relative Cerebral Blood Flow Can Assist in Triage of Stroke Patients to Reperfusion Therapies

Diffusion MRI of the brain, axial section, showing left cerebrovascular ischemia 5 hours after a stroke in a 60 year oldmale patient. The FLAIR (left), diffusion (middle), and app arent diffusion coefficient (ADC; right) sequence images show an ischemic area around the left middle cerebral artery, corresponding to an interruption in blood circulation.
Study confirms a high correlation and good agreement between infarct volumes measured using acute diffusion-weighted imaging and relative cerebral blood flow derived from temporally and volumetrically matched perfusion data.

Relative cerebral blood flow (rCBF) can accurately differentiate between small brain infarcts and large infarcts, with diagnostic performance that is sufficient for clinical triage of patients to reperfusion therapies, according to study results published in Stroke.

While thrombectomy is an effective treatment for acute ischemic stroke, it is not indicated in patients with large infarct cores who have an increased risk for treatment complications. Diffusion-weighted imaging (DWI) is the reference standard for delineating the infarct core. Although DWI is more accurate for detection of irreversible tissue injury, computed tomography (CT) is faster and easier to access, and thus is more widely used. CT perfusion can be used to estimate the infarct core, and rCBF is accepted as the optimal perfusion-derived parameter for predicting infarction.

Related Articles

In this study, the researchers assessed whether rCBF is sufficiently accurate to distinguish between large and small infarct volumes, thus aiding in identifying appropriate candidates for reperfusion therapy.

Of 193 baseline MRI scans from 193 patients enrolled in the DEFUSE 2 (Diffusion and Perfusion Imaging Evaluation for Understanding Stroke Evolution) and SENSE 3 (Sensitivity Encoding) studies who were screened, 119 scans (from 119 patients) with DWI and perfusion imaging within 24 hours of onset were analyzed.

Twelve closely spaced rCBF thresholds were used: 0.20, 0.22, 0.25, 0.27, 0.30, 0.32, 0.34, 0.36, 0.38, 0.40, 0.42, and 0.44. For each rCBF threshold, the agreement and correlation between infarct core volumes estimated using DWI and rCBF were assessed.

The rCBF threshold of 0.32 yielded the best spatial and volumetric agreement between infarct core estimates obtained using rCBF and DWI. This threshold yielded the smallest mean absolute volume difference of 14.7 mL, with no statistically significant difference between the DWI and rCBF volumes (P =.84). All rCBF thresholds below 0.30 and above 0.34 yielded a significant difference (P <.05) between DWI and rCBF volumes. There was a strong positive correlation between DWI and CBF volumes across all rCBF thresholds tested.

The optimal rCBF threshold did not significantly differ between patients treated in the early window vs the late window. Researchers identified the number patients incorrectly identified at each rCBF threshold with 70 mL as the volume limit for triage to thrombectomy. Of 119 patients, 110 (92.4%) received correct triage management using the rCBF-derived infarct volume at the 0.32 threshold.

The study had several limitations including the major limitation of CBF imaging that provides a snapshot of hemodynamic derangement, the limited number of patients seen during the hyperacute time window for meaningful analysis of the optimal rCBF threshold, and that DWI abnormality to measure the infarct core can partially reverse with reperfusion.

“This study confirms a high correlation and good agreement between infarct volumes measured using acute DWI and rCBF derived from temporally and volumetrically matched perfusion data. rCBF could reliably distinguish patients with small infarcts from those with large infarcts, hence its diagnostic performance is sufficient for clinical triage of patients to reperfusion therapies,” concluded the researchers.


Amukotuwa S, Straka M, Aksoy D, et al. Cerebral blood flow predicts the infarct core: new insights from contemporaneous diffusion and perfusion imaging. Stroke. 2019;50(10):2783-2789.