Neurovascular Diseases

Pathophysiology

Normal cerebral blood flow (CBF) is 50 to 55 mL/100 g brain tissue/min. Cerebral autoregulation responds to a fall in cerebral perfusion pressure (CPP) with vasodilatation and recruitment of collateral vessels, thus increasing cerebral blood volume (CBV) and reducing resistance, in order to maintain CBF. The average time a blood cell remains with a particular volume of tissue rises due to vasodilatation and collateral flow, resulting in a prolonged mean transit time (MTT) and thereby allowing improved oxygen delivery. After the vessels are fully dilated the autoregulatory system cannot properly respond to any further reduction in CPP and therefore CBF starts to decline. Oxygen extraction goes up to compensate, but once this is maximal any further fall in CBF causes cellular dysfunction. The loss of normal neuronal electrical function occurs when CBF falls to 15 to 20 mL/100 g/min. However, this may be reversible, depending on the severity and duration of the ischaemia, such that irreversible infarction is likely to occur within minutes if the CBF less than 10, but moderate ischaemia (10 to 20) may be reversible for a few hours. At levels of CBF less than 10, hypoxaemia leads to failure of the adenosine triphosphate (ATP)-driven cell integrity systems (glutamate, N -methyl-D-aspartate [NMDA], Na ⁺ /K ⁺ ), resulting in cell depolarisation and influx of Na ⁺ and water. Cellular swelling and cell death occur (cytotoxic oedema). In time, structural breakdown of the blood–brain barrier occurs due to ischaemic damage to capillary endothelium. Leakage of intravascular fluid and protein into the extracellular space and later net influx of water to the infarcted area cause vasogenic oedema. However, CBV generally remains preserved in infarction, unless there is a profound reduction in CPP, where it has been postulated that microvascular collapse due to inability of the vessels to remain patent may eventually result in a reduction in CBV.

Following recanalisation of the occluded vessel, either spontaneously (33% within 48 hours) or following treatment, the ischaemic region becomes reperfused. This will occur with both viable and non-viable tissue. Indeed, a state of ‘post-ischaemic hyperperfusion’ ensues where persistently vasodilated vessels result in an elevated CBV. CBF is also elevated.

The Penumbra Model

Following a thromboembolic cerebral arterial occlusion, the decline in regional CBF in the affected brain parenchyma is not uniform. The accepted model, validated in animals and humans, is centred upon an infarct core with very low CBF and cell depolarisation. A peripheral zone—the penumbra—has moderately diminished CBF, resulting in loss of electrical function but preserved cell integrity. The duration of ischaemia in the penumbra is critical, and strategies to recanalise the vessel and restore normal CBF are likely to convey the greatest benefit here. However, failure or, more crucially, a delay in achieving this may lead to progression to infarction, especially as this tissue is poorly autoregulated and more vulnerable. Surrounding the penumbra is a zone of benign oligaemia. Here CBF is only mildly impaired and tissue is likely to survive.

Stroke Classification

The most commonly used classification system of ischaemic stroke (TOAST—trial of ORG 10172 in acute stroke treatment) discriminates between large vessel thromboembolic, cardioembolic, small vessel and ‘other’ aetiologies. Precise allocation into these subtypes is sometimes difficult and strokes are not infrequently undetermined. A few points to consider are:

• Deep white matter infarcts are typically small vessel in nature but can result from emboli originating from large vessel atheroma or from a cardiac source.

• Middle cerebral artery (MCA) territory infarcts can arise from emboli from the heart or carotid artery, or from in situ thrombosis in the MCA.

• Small peripheral infarcts in a vascular territory are usually embolic but the source is not always clear (i.e. cardiac vs. carotid vs. MCA) ( Fig. 56.1 ).

  

• Peripheral infarcts involving multiple vascular territories must be from a proximal source and most likely the heart.

• The basilar artery supplies the posterior cerebral arteries (PCAs) unless the posterior communicating artery(s) is/are large, in which case emboli from the carotid circulation may enter their territory. Brainstem infarcts commonly result from occlusion of short perforating vessels. A combination of infratentorial, thalamic and occipital infarcts suggests an occlusion of distal basilar artery, or ‘top of the basilar’ syndrome ( Fig. 56.2 ).

  

Causes

Borderzone Infarction

Also known as watershed ischaemia, this occurs at the boundaries of the major vessel territories—superficially between the leptomeningeal collaterals of the MCA and ACA, which also extend into the corona radiata deep to the superior frontal sulcus, and those of the MCA and PCA. In the deep white matter of the inferior corona radiata and external capsules lies the deep border zone between the cortical branches and deep M1 perforators of the MCA ( Fig. 56.3 ).

Postulated mechanisms include local (e.g. carotid stenosis) and global (e.g. cardiac insufficiency) hypoperfusion, but embolic infarcts at these sites can also occur.

Borderzone ischaemia in the posterior fossa is uncommon but usually occurs between the superior cerebellar artery (SCA) and posterior inferior cerebellar artery (PICA) territories, and occasionally between the SCA, PICA and anterior inferior cerebellar artery (AICA) territories.

Global Hypoxic–Ischaemic Injury

Inadequate oxygen supply to the entire brain can be the consequence of severe hypotension or impaired blood oxygenation. Global hypoperfusion can result in watershed infarcts as described above, but profound hypoxia can also cause symmetric ischaemia in the basal ganglia, thalami and hippocampal formations. Anoxia due to defective blood oxygenation such as in carbon monoxide poisoning tends to cause infarcts in sensitive regions, typically, as in this case within the globus pallidus ( Fig. 56.4 ).

Imaging Strategies and Goals in Acute Stroke

Brain imaging is vital in the paradigm of acute stroke assessment. Previously, the only licensed therapy for vessel recanalisation in the acute period involved the administration of intravenous (IV) thrombolytic agents—mainly recombinant tissue plasminogen activator (r-tPA). This was derived from key studies (NINDS, ECASS-3 and SITS-MOST), which demonstrated significant improvements in the degree of disability at 3 months in patients with ischaemic stroke treated with IV r-tPA provided they were within 4.5 hours of onset, had sustained infarcts of less than one-third of the MCA territory and were less than 80 years old. Around 12% of all acute stroke presentations in the UK currently receive IV thrombolysis. Traditionally, imaging has been directed at these criteria, employing cranial non-enhanced computed tomography (NECT).

The year 2015 heralded a significant turning point in the treatment options for patients presenting with acute stroke syndromes secondary to a large vessel occlusion (LVO), specifically for lesions affecting the intracranial ICA and M1 segment of the MCA. Traditionally, these patients present with major clinical syndromes and respond poorly to medical therapy. Advances in endovascular clot retrieval devices and the publication of a number of landmark randomised trials (MR CLEAN) and a meta-analysis of these trials (HERMES) demonstrated that mechanical thrombectomy provided a clear sustained benefit over standard medical care in reducing disability at 90 days in patients who present with acute anterior circulation stroke secondary to LVO.

Acute stroke management guidelines have been updated across the world to reflect this literature. In the UK, new guidance recommends mechanical thrombectomy should be provided to patients with a moderate or severe acute stroke syndrome (NIHSS >5), occlusion of the intracranial ICA or proximal MCA and who have a pre-stroke disability score (Modified Rankin Scale-mRS) of less than 3. If eligible they should receive treatment with IV r-tPA within 4.5 hours of stroke onset and should undergo arterial puncture within 6 hours. The literature suggests that up to 10% of patients presenting with acute stroke may be eligible for mechanical thrombectomy. Patients fulfilling the above criteria for thrombectomy should in addition to the NECT at the time of presentation have a CT angiogram from the arch of the aorta to the intracranial circulation to assess for an LVO. A proportion of LVOs will be missed if the ‘hyperdense vessel sign’ is relied on as a surrogate marker of the LVO.

Objectives of Non-Enhanced Cranial Computed Tomography in Acute Stroke

• To exclude haemorrhage and allow administration of aspirin therapy.

• To exclude an alternative cause of the fixed neurological deficit. Around 30% of patients presenting with a stroke-like episode have a non-vascular cause.

• To assess infarct volume and Alberta Stroke Program Early CT Score (ASPECTS).

• To assess for a hyperdense vessel.

• To establish that the infarct corroborates the clinical timeline and does not obviously appear subacute.

Hyperacute Infarct Imaging Signs on Computed Tomography

A dense artery is the earliest detectable change on computed tomography (CT). As it is caused by fresh thrombus occluding the vessel it can be seen at the onset of the ictus. Thrombus may rapidly disperse, so this sign is not always present. When found in the proximal MCA or terminal ICA, it correlates with large infarcts and very poor outcomes although it has a better prognosis if limited to an MCA branch within the Sylvian fissure (the Sylvian fissure ‘dot’ sign) ( Fig. 56.5 ). MCA calcification can mimic this sign but is often bilateral. The basilar artery may also appear dense in the case of posterior circulation infarcts, particularly the ‘top of basilar’ syndrome. The early parenchymal signs on CT are reduced grey matter density and brain swelling, manifest as effacement of sulci ( Fig. 56.6 ). These changes are traditionally thought to reflect cytotoxic oedema, which reduces the Hounsfield number of grey matter so it is indistinguishable from adjacent white matter. In early MCA infarcts this causes a reduction in clarity of the margins of the lentiform nucleus and cortex, particularly in the insula. Hypodensity on early CT examinations affecting more than 50% of the MCA territory is associated with a high mortality rate, and IV thrombolysis is contraindicated when more than one-third of the MCA territory is involved. However, infarct size evaluation is notoriously difficult in the acute phase, due to the lack of convincing parenchymal changes in 50% to 60% of NECT within 2 hours. The sensitivity of CT for infarcts has been reported to be only 30% at 3 hours and 60% at 24 hours. These difficulties have led to the development of the ASPECTS. ASPECTS can be used to predict outcome and risk of post-thrombolysis haemorrhage. It correlates well with diffusion-weighted imaging (DWI) findings at presentation and facilitates more accurate interpretation of emergency CT by non-experts. Even in patients not suitable for thrombolysis it seems intuitive that a methodical approach, such as ASPECTS is likely to increase accuracy and reliability of CT interpretation, at least for supratentorial events ( Fig. 56.7 ). The sensitivity to subtle grey matter low attenuation is enhanced using the ‘stroke window’ setting when reviewing images (window width = 35/window level = 35).