Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Noncontrast computed tomography (CT) is the standard diagnostic modality for acute stroke patients. It reliably differentiates hemorrhagic from ischemic stroke, enables rapid thrombolysis, and thereby improves stroke recovery.
On noncontrast CT, different types of “early ischemic changes” can be found: hypodensity, isolated cortical swelling, and hyperdense arteries. Patients with extensive hypodensity do not appear to benefit from thrombolysis, whereas isolated cortical swelling or hyperdense arteries should encourage recanalization therapies.
CT angiography (CTA) is a fast and reliable tool to diagnose and grade intra- and extracranial occlusive disease. Patients with a large thrombus burden on CTA might benefit from endovascular recanalization.
In patients with posterior circulation stroke, CTA enables rapid diagnosis of basilar thrombosis.
CTA improves prediction of irreversible brain tissue injury using a low contrast and window level.
CT perfusion (CTP) provides additional information about extent of irreversible brain injury as well as extent of salvageable tissue that is not possible with noncontrast CT and CTA.
Whole-brain CTP can also allow dynamic CTA acquisition, which provides valuable insights into collateral flow in acute ischemic stroke.
We would like to acknowledge Charles A. Jungreis and Steven Goldstein for their work on previous editions of this chapter.
This chapter studies the clinical efficacy of computed tomography (CT) in patients with acute ischemic or hemorrhagic stroke. We will discuss noncontrast CT (NCT) as well as CT angiography (CTA) and CT perfusion (CTP) imaging for acute stroke. Because of the differing prognosis between anterior and posterior circulation ischemic strokes, we will separately discuss CT imaging for these. In theory, CT, like magnetic resonance imaging (MRI), can be clinically effective in patients with acute stroke on five different levels: (1) technical capacity, (2) diagnostic accuracy, (3) diagnostic impact, (4) therapeutic impact, and (5) patient outcome or prognostic impact.
We will focus on the use of CT in the acute stroke setting. Whenever applicable, we will compare CT with MRI—the other important acute stroke imaging modality discussed in Chapter 48 .
A potential advantage of CT over MRI is its wide availability and excellent feasibility. Even with older-generation scanners a plain CT scan can be performed within minutes. Performing a CTA or CTP requires a spiral CT scanner, the application of contrast media, and processing afterward. Each examination will usually only require an extra 5 minutes’ scan time. Recently, CT scanners with up to 320 detector rows that cover a tissue volume of 16 cm thickness have become available. These scanners can examine the whole brain in less than 1 second and can repeatedly image the entire cerebral circulation, allowing time-resolved images of the brain vessels and whole-brain CTP.
The main downsides of CT brain imaging are radiation and the use of iodinated contrast agents. The radiation dose per CT scan is in the range of 1–3 mSv depending on the scanner type. Therefore, repeating a CT to follow-up brain and vessel pathology, and in particular the repetition of CTP, can be problematic. For brain imaging the lenses should be protected. The risk of kidney function failure or thyrotoxic crisis caused by iodinated contrast agents is relatively low and should not delay imaging but requires tests of kidney and thyroid gland function. Another disadvantage of unenhanced CT is that automatically measured volumes of pathologic tissue, such as can be performed on CTP or apparent diffusion coefficient maps, are not available.
Acute brain ischemia below the cerebral blood flow (CBF) threshold of 20–30 mL/100 g/min leads to loss of neurologic function, cell membrane dysfunction with cellular edema (so-called cytotoxic edema), and subsequent shrinkage of the extracellular space. This type of edema is potentially reversible and can be visualized by MRI with diffusion-weighted imaging (DWI). Acute severe brain ischemia with CBF values below 10 mL/100 g/min causes immediate net water uptake into gray matter. , This so-called ionic edema characterizes brain tissue destined for tissue necrosis, even with early reperfusion. Only this net uptake of water into brain tissue causes a decrease in x-ray attenuation on CT. The decrease in x-ray attenuation is linearly and indirectly correlated to the amount of water uptake. A 1% increase in tissue water causes a decrease of approximately 2–4 Hounsfield units (HU) in x-ray attenuation that can be detected by the human eye. CT is thus a very specific method for depicting irreversible brain infarction, and once infarction is apparent on CT, it is irreversible.
Early changes visible on head CT during ischemia are often summarized as “early ischemic changes” (EIC). We should, however, distinguish between at least three different kinds of EIC with different pathophysiologic and diagnostic relevance:
Reduced x-ray attenuation of gray matter is the common cause of CT signs such as “loss of the insular or cortical ribbon,” “obscuration of the lentiform nucleus,” reduction in gray matter–white matter contrast, or “hypodensity.” These phenomena are all consequences of ionic cerebral edema and thus represent ischemic infarction on CT ( Fig. 47.1 ).
Isolated brain tissue swelling without reduced x-ray attenuation ( Fig. 47.2 ): this phenomenon has been extensively studied recently and is likely caused by compensatory vasodilation with an increase of cerebral blood volume (CBV). It represents tissue at risk of infarction that might be salvaged with reperfusion, i.e., ischemic penumbra. It was observed in 10%–20% of ischemic brain regions. The presence of isolated cortical swelling on CT—even if very extensive—should NOT prevent stroke neurologists from attempting urgent recanalization.
The hyperdense artery sign is highly specific for the presence of an intra-arterial thrombus, but its detection depends on the thrombus’ hematocrit value, which determines its x-ray attenuation. Recently, it has been shown that accuracy of thrombus detection can be markedly improved using thin-slice (1.25 or 2.5 mm) reconstructions. Using these, thrombus length can be accurately measured. Thrombi with a length of greater than 8 mm may have no chance to be fully lysed by intravenous (IV) thrombolytics. The term EIC should not be used for the hyperdense artery sign because arterial obstruction by an intraluminal thrombus can be fully compensated by collateral flow (see Fig. 47.1A ). Additionally, thrombectomy is a highly effective therapy for long clots.
In the first 6 hours after ischemic stroke, two of three stroke patients will develop ionic edema that can be detected by NCT. Poor sensitivity of NCT is often blamed for the fact that ischemic (i.e., ionic) edema cannot be diagnosed early in many patients with ischemic stroke. Stroke symptoms occur at CBF reduction to 20–30 mL/100 g/min, which is well above the accepted threshold for ionic edema. We can thus assume that around one-third of ischemic stroke patients have not yet developed relevant ionic edema (i.e., no irreversible damage) and may have an excellent prognosis if reperfusion is achieved. X-ray hypoattenuation on early NCT, however, is subtle and hard to detect without training (see Fig. 47.1B ). Interrater reliability for ischemic tissue hypoattenuation varies between a kappa value of 0.4 and 0.6. The “sensitivity” of NCT for infarction on follow-up imaging varies between 20% and 87% depending on image quality, experience, and training. Using systematic scores like the Alberta Stroke Program Early CT Score (ASPECTS) facilitates detection of x-ray hypoattenuation and improves reliability and sensitivity. Once a hypoattenuating area has been detected, it is highly predictive of subsequent infarction.
The diagnostic impact of stroke imaging refers to the proportion of patients in whom the specific diagnosis of stroke type relies on imaging. NCT has a huge diagnostic impact simply by differentiating ischemic from hemorrhagic stroke. In addition, the extent of hypoattenuation on NCT is an important positive predictor of thrombolysis-induced brain hemorrhage, which is the most feared complication of thrombolysis. MR DWI is highly sensitive for ischemic brain tissue, even above the CBF level of the penumbra, and indicates brain tissue at high risk, if not already irreversibly injured, whereas hypoattenuation on NCT depicts irreversible tissue damage with high specificity. High signal intensity on DWI may allow assessment of ischemic brain patterns and thus the cause of stroke early on and signals an increased early risk of recurrent stroke in patients with transient ischemic attacks (TIAs).
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here