Acute cerebral perfusion imaging should be part of the routine evaluation of patients with acute ischemic stroke secondary to large vessel occlusion presenting beyond 6 hours from time last known well, and may help identify those who may benefit from endovascular perfusion therapy. However, it is important to understand the limitations of automated CT perfusion (CTP), which are detailed in a review published in Neurology.
The clinical benefit of endovascular reperfusion therapy was impressive in 2 randomized trials (DAWN and DEFUSE 3) for patients with acute ischemic stroke with larger vessel occlusion who presented beyond 6 hours from last known well and were deemed to be good candidates based on cerebral perfusion imaging. The capability to perform and interpret perfusion imaging, either CTP or magnetic resonance perfusion-weighted imaging, should be available in every center treating patients with acute stroke.
A few commercially available, automated software packages, such as RAPID automated perfusion postprocessing software, can detect the presence of irreversible brain tissue injury, based on cerebral blood flow compared with that seen in contralateral perfused tissue by CTP or magnetic resonance perfusion, and assist in selecting appropriate candidates for endovascular therapy.
Simply put, CTP consists of a sequence of images obtained during the wash-in and the wash-out of an intravenous iodinated contrast agent. This is followed by transferring the source images to the workstation with the dedicated postprocessing software, which provides calculation of hemodynamic parameters. The automated assessment overcomes possible observer variability associated with subjective inspection.
There are disagreements about the quantitative thresholds between the ischemic core (the critically hypoperfused tissue with low cerebral blood flow) and the ischemic penumbra (the potentially salvageable tissue that may recover if reperfusion is achieved), and different software packages recommend slightly different thresholds. The automated software RAPID generated specific thresholds to identify good candidates for delayed endovascular: relative cerebral blood flow <30% and Tmax (time from the start of the scan until the maximum intensity of contrast bolus arrives at each voxel) >6 seconds.
Although perfusion imaging can be very valuable, there are several diagnostic pitfalls and challenges, and the review details the common technical and clinical pitfalls, along with practical pearls.
Technical pitfalls refer to the image quality from a technical standpoint and include several important challenges:
- Patient motion: as the quality of the image may be degraded with even slight motion, it is important to always check for patient motion.
- Poor contrast bolus: contrast arrival may be delayed in patients with large vessel occlusions, and may also be insufficient with inadequate intravenous access. Confirming the arterial input and venous output function curves of a CTP study can point to an adequate baseline, upslope, and downslope without truncation.
- Brain coverage in CTP: it is important to know the available CT scanners and whether they provide coverage of the entire brain in 1 or 2 slabs.
- Excessive radiation for CTP: as CTP requires repeat scanning of the head, it results in relatively high radiation exposure. Furthermore, the radiation dose to other organs may be increased in cases of erroneous CTP protocols. For these reasons, it is vital to optimize CTP protocols to reduce the radiation dose.
Clinical challenges to consider in routine practice:
- Underestimation of the ischemic core: calculation of ischemic core is based on predefined thresholds, which were derived from large data sets aiming to be more inclusive in patient selection and therefore allow more underestimation of core volumes. For that reason, it is advised not to focus on automated CTP outputs, but always look at the entire study, including noncontrasted CT and CT angiogram.
- Calculating core volumes using a set threshold: as CTP is a snapshot in time, core volume overestimation is possible in cases of ultraearly reperfusion. It is recommended to look at additional thresholds to assess the total ischemic core and to keep in mind that optimal CTP thresholds may vary from patient to patient.
- Penumbral misclassification: misclassification of the ischemic penumbra may be secondary to technical or flow-related patterns. As the calculation of penumbral volume produced by the software may include artifacts, this may cause overestimation of the penumbral volume. It is advisable to recognize the possibility for misclassification of the penumbra in patients with carotid stenosis, low cardiac output, or cardiac arrhythmia.
- Posterior fossa and lacunar stroke: limited information is available on the appropriate thresholds recommended for perfusion imaging of the posterior circulation. In addition, CTP has a low sensitivity for detection of lacunar strokes. It is important to understand the limitations of CTP in these cases and consider completing MRI, which is a better alternative.
- Intervendor variability in CTP results: as there is no standardization between CTP vendor software and postprocessing techniques, variations in calculated core and mismatch volumes are an important challenge.
- Stroke mimics: seizures, hypertensive encephalopathy, hemiplegic migraines with vascular dysregulation, or variations in vascular anatomy can be falsely classified as ischemic penumbra. For that reason, not every CTP abnormality represents a stroke.
Magnetic resonance perfusion is less quantitative compared with CTP and requires the use of echo planar imaging, which can have multiple artifacts. Furthermore, MRI is not as commonly available in urgent cases.
Lead author Achala Vagal, MD, professor of radiology at the University of Cincinnati Medical Center, Ohio, told Neurology Advisor, “CTP is an important tool in the triage and treatment decision making in acute ischemic stroke. However, like any other technique, it is imperative to understand its limitations, especially when these decisions have to be made quickly.” She goes on to add that “automated CTP output maps are now available within minutes on our PACS, phones and emails; all the more reason to be cognizant of the pitfalls leading to inaccurate estimations of core and penumbra. A quick checklist can be very helpful.”
Vagal A, Wintermark M, Nael K, et al. Automated CT perfusion imaging for acute ischemic stroke: pearls and pitfalls for real-world use. Neurology. 2019;93(20):888-898