Imaging in Stroke: The More You Look, The More You See

Cerebral imaging has both humbled and enlightened stroke clinicians over the last 30 years. Imaging techniques are used to assist in diagnosis, guide intervention, and facilitate research. Stroke is not the only cause of sudden focal neurologic deficit, and imaging must help distinguish tumors, extraaxial lesions, migraines, and seizures. Clinical stroke research has become dependent on imaging in an effort to sharpen the understanding of the dynamic processes involved. Clinicians have struggled to define cerebral function and blood flow, relying on positron emission tomography, single-photon emission computed tomography, or Xenon imaging to provide data. More recently MR has promised, and may yet deliver, rapid physiologic data with diffusion and perfusion imaging.

Clinicians have long used the arbitrarily defined terms stroke and transient ischemic attack (TIA) to refer to the sudden loss of neurologic function from a vascular mechanism. Symptoms resolving in less than 24 hours have been referred to as TIA. The underlying tissue physiology of ischemia, whether reversible injury or infarction, does not always correspond to the clinical terms of TIA or stroke. Patients with TIA are commonly found to have areas of tissue infarction when imaged carefully. Many stroke clinicians have come to favor a 1-hour time limit for TIAs, improving the correlation of clinical and physiologic terms. Neuroimaging has become tightly integrated in the study and management of cerebrovascular disease.

In the decade since the first MR diffusion imaging of stroke patients, the technique has evolved and become familiar to those involved in acute intervention. Diffusion-weighted imaging has come to be a sensitive early marker of infarction, and reversible diffusion abnormalities have been far less common clinically than in animal stroke models. Territories with perfusion and diffusion mismatch may define tissue at risk for infarction, but with potential for recovery. An alternate strategy with CT technology uses rapid CT for dynamic perfusion imaging, with similar goals in mind. It is hoped that acute imaging can better guide interventions such as intravenous or intraarterial thrombolysis. While most attention has been focused on acute intervention, it must be emphasized that the vast majority of patients are not seen in an appropriate timeframe for acute therapies. Less than 10% of all stroke patients are evaluated less than 3–6 hours following symptom onset, severely limiting the application of acute therapies.

The vast majority of stroke patients are seen in the subacute time frame, where diagnostic imaging also plays a vital role in directing management. In this issue of the AJNR, Augustin et al (page 1596) report on the use of diffusion-weighted imaging in subacute stroke. They observe that diffusion-weighted imaging adds sensitivity to the standard MR evaluation, allowing identification of lesions otherwise not detected. The specificity of diffusion imaging for recent infarction also increases the ability to detect new lesions in a background of chronic changes.

Secondary stroke prevention is a major component of management in all patients, and represents the major focus in the subacute group. While experience would suggest that small vessel disease leads to 20% of ischemic stroke cases, another 20% are caused by large vessel carotid and intracranial disease, and 60% of cases derive from thromboembolic events, decisions for each patient should relate to a specific demonstration of stroke etiology. The evaluation is often directed by the diagnostic imaging. Small, deep infarcts typical of small vessel lacunar disease are usually easily identified, and imaging adds greatly to diagnostic certainty compared to clinical impression alone. While these may not always be related to intrinsic small vessel disease associated with hypertension and diabetes, the odds are such that embolism becomes a much less likely cause. Peripheral cortical infarcts imply an embolic mechanism, and if in the anterior circulation, should lead to evaluation of the cervical carotid. In these cases, more information is helpful. If diffusion imaging reveals a clinically silent subacute infarct in the cerebellar hemisphere, then interpretation of a 70% stenosis in the cervical internal carotid is radically altered. What might be interpreted as symptomatic carotid disease now becomes multiple emboli in different territories and leads to a search for cardiac sources, and therapy may change from surgical intervention to systemic anticoagulation.

It is interesting that despite the sensitivity of current imaging, it is not possible to identify an appropriate lesion in all patients with acute stroke deficits. Optimal stroke care requires collaboration between the clinician and radiologist, for we are clearly far from total understanding of the process. Diffusion-weighted imaging makes an important contribution to stroke management, even in the subacute timeframe, and should become widely applied. As improvements in technology allow further investigation, the more information we will have to be considered and applied to patient care.