Stroke Imaging Comes of Age: Updated Stroke Definition Embraces Modern Neuroimaging and Highlights Need for Standardized Imaging Protocols

Stroke mortality in the US has declined in the last few decades largely due to the use of antithrombotic therapy and lower blood pressures in the population.(1) Despite these advances, stroke remains the number 4 cause of death in the United States; 6.8 million Americans ≥ 20 years old have had a stroke, and an additional 4 million individuals are projected to have a stroke by the year 2030.(2) Treatment strategies for stroke vary depending on stroke subtype, but are fairly standardized across the more than 900 Joint Commission--certified Primary Stroke Centers in the US. However, to date a clear consensus on the diagnostic criteria for stroke and its subtypes has been difficult to establish despite repeated efforts.

In 1994, Goldstein and Matchar proposed a method of rapid and accurate clinical diagnosis of stroke based on physical findings.(3) This proposal attempted to overcome limitations in CT-based detection methods available at the time, and resulted in the broad definition of stroke as an acute onset of focal neurological deficits with no pre-existing trauma, subject to verification by repeated CT or MRI scans, lumber puncture, angiography, cardiac diagnostic studies, clinical course, and autopsy for final diagnosis.

This basic definition was updated in 2005 to include diagnostic parameters that are important not only for stroke specialists but also for first responders.4 However, this update still relied heavily on physical findings. For instance, 3 physical findings, facial paresis, arm drift, and abnormal speech, which were derived from the National Institutes of Health Stroke Scale (NIHHS) and incorporated into the Cincinnati Prehospital Stroke Scale, predicted strokes with a sensitivity of 100% and specificity of 88%, and were thus recommended as rapid diagnostic tests for both field and hospital settings.(4)

These reports and others from the same era emphasized physical stroke findings partly because of limitations with, and partly because of lack of uniform access to, CT and MRI imaging facilities at that time. However, times have changed, and with those changes have come a need for a revised definition of stroke. The AHA/ASA Expert Consensus Document “An Updated Definition of Stroke for the 21st Century”5 delivers this much-needed update to our understanding of stroke, and sets a new standard for management of stroke patients as well as for clinical and public health research.

Advances in CT- and MR-based imaging techniques have made it possible to localize regions of brain infarction and hemorrhage at high spatial resolution. Specific vessels that supply the affected regions can be inspected for occlusions, malformations, or aneurysms using duplex Doppler and transcranial Doppler ultrasound for the neck and head, respectively, using CT or MR angiography, or using catheter angiography. Furthermore, regions of the so-called penumbra, which are ischemic but not yet permanently damaged and thus salvageable by intensive reperfusion therapy, can now be visualized by multimodal MRI imaging, which combines MR angiography, diffusion-weighted imaging (DWI), fluid attenuated inversion recovery (FLAIR), and perfusion-weighted MRI (PWI), or by mismatches in CT hypodensity, CT angiography, and CT perfusion images.

An important consequence of these new imaging techniques is that smaller lesions such as silent infarcts and microhemorrhages can now be readily detected, bringing up the issue of whether or not these lesions should be counted as strokes. Lumping silent infarcts along with strokes would increase the number of stroke diagnoses from roughly 3% to 15% to 20% of the population, albeit reducing the overall rate of stroke mortality. Also the number of these detectable lesions will likely continue to expand over time as clinical imaging techniques improve. Yet silent infarcts may increase the risk of mild cognitive impairment, vascular dementia, and Alzheimer’s disease, and treating patients with silent infarcts may prevent future frank strokes. Therefore, the Consensus document recommends that patients with silent infarctions and microhemorrhages be treated with measures of primary stroke prevention and that future studies address the safety and efficacy of secondary prevention measures in these patients.

Improved imaging techniques have also shed light on transient ischemic attacks (TIAs), which have traditionally been defined as clinical findings persisting for less than 24 hours. With advanced imaging, it is now evident that some clinically transient ischemia’s show evidence of brain infarcts. Likewise, some protracted ischemia’s that are currently defined as strokes do not show any permanent brain infarction upon imaging. The Consensus document therefore calls for a new functional definition of TIAs based on imaging studies rather than clinical findings alone. Further studies are needed to carefully characterize what imaging findings correspond to TIAs and to verify such findings in autopsy specimens; however, available data suggest that the 24-hour mark is still a reasonable cutoff for TIA definition.

Caution should be exercised in abandoning clinical stroke diagnostics in favor of imaging-based techniques, since high-resolution MRI and CT scanners are not readily available in field locations where emergency medical personnel operate, in many US hospitals, and in most hospitals in the developing world. Clinical wisdom developed and validated over decades is still the most viable or only option in such cases. However, more and more hospitals are acquiring state-of-the-art imaging facilities, and telemedicine is becoming a reality in many remote locations such that most stroke patients now have access to imaging-based evaluations. Also of note, the Consensus document does not propose delaying treatment until all imaging data are available. Rather, similar to patients with acute coronary syndrome, the blanket term ”acute cerebrovascular syndrome” may be used for patients presenting with cerebral infarction, TIA, or hemorrhage, and immediate treatments are recommended to minimize permanent brain tissue damage while awaiting imaging results.

In summary the Consensus statement provides a much-needed update to our terminology for stroke and takes into account recent advances in medical imaging. The recommendations set forth in the Consensus statement set new clinical benchmarks for the diagnosis and standardized treatment of patients with various subsets of stroke. Additionally, the updated definitions of stroke and TIAs constitute a call for standardization of imaging protocols that may be used for patient care, for clinical trials, and for future clinical and epidemiological studies on the prevention and treatment of stroke.(5)