The kinetic behavior of diffusible perfusion tracers such as oxygen 15 [¹⁵O]H₂O is crucially different from tracers that get trapped, such as the SPECT agent technetium 99m [^99mTc]HMPAO. This difference warrants a completely different acquisition protocol. Since [^99mTc]HMPAO uptake does not change following injection, imaging can be performed as a single scan from minutes to hours following injection. In contrast, [¹⁵O]H₂O is washed in and out of the tissue within minutes. Together with the short physical half-life of [¹⁵O], this requires imaging to be performed within minutes following injection if one uses a bolus injection. An interesting alternative method for certain applications is the use of a constant infusion protocol, which allows the continuous monitoring of blood flow (1). A detailed description of the methods used for [¹⁵O]H₂O imaging is given in the appendix at the end of the chapter.

Following inhalation as a bolus, [¹⁵O]CO is irreversibly bound to hemoglobin. Its distribution is therefore restricted to intravascular space. Thus, the ratio of the [¹⁵O]CO concentration in the brain and in blood directly yields the CBV.

Stenoses of the larger arteries feeding the brain are common. An important clinical aspect in this regard concerns the hemodynamic significance. There are two mechanisms by which artherosclerotic plaques may lead to symptoms such as transient ischemic attacks (TIAs) or stroke. Emboli may break lose from the plaques and cause the symptoms. Another possibility is that the stenosis is severe enough to lead to compromised cerebral perfusion in areas fed by the affected artery. The cerebral vascular response to reduced perfusion is complex. As a first response to decreased perfusion, the brain increases the oxygen extraction fraction. Another response is the dilatation of the vessels in an attempt to reduce vascular resistance and counteract reductions in perfusion. When the potential of these mechanisms is fully exploited and the perfusion still decreases, the patient becomes symptomatic. For instance, if the perfusion is just sufficient with maximum oxygen extraction and fully dilated vessels, any further stress will lead to decompensation. Such patients may benefit from a revascularization procedure such as endarterectomy or extracranial to intracranial (EC-IC) bypass surgery. The improvement of cerebral perfusion following revascularization has been demonstrated in several studies (2,3,4,5,6,7). An example of a patient with a severe stenosis of the right internal carotid artery is shown in Figure 27.1.

There exist several methods and parameters for evaluating patients with chronic cerebrovascular disease. Based on the physiological regulation of cerebral blood flow, the assessment of blood volume and perfusion is useful. This is illustrated in Figure 27.2A. The images on the left sketch the vascular autoregulation that occurs in cerebrovascular disease. The most common site for stenoses is the carotid bifurcation. The regulatory steps include increase of the oxygen extraction and dilatation of the vessels to reduce vascular resistance. The latter step is equivalent to an increase in cerebral blood volume (CBV). One method of evaluating the hemodynamic situation is indeed based on the measurement of CBV. The CBV/CBF ratio is an especially useful measure. The critical situation is reached when the vessels are fully dilated and CBF continuous to decrease. This situation is reflected by an increasing CBV/CBF ratio. Alternatively, one can measure hemodynamic behavior during induced vasodilatation. A strong stimulus for vasodilatation is an increase in arterial CO₂ (PaCO₂). Hypercapnia can be induced by inhalation of an CO₂-enriched air mixture or by the administration of acetazolamide (Diamox).

In healthy subjects, the reduced vascular resistance due to induced vasodilatation leads to an increase in CBF. The ratio of CBF during hypercapnia and at baseline is often referred to as the cerebral perfusion reserve. The situation with a stenosis is illustrated at the bottom of Figure 27.2A. rCBF is normal at baseline due to chronic vasodilation on the compromised side. Blood flow images will reveal symmetric perfusion in both hemispheres. Inducing hypercapnia will now dilate the vessels on the healthy side but not on the affected side or to a lesser degree, as they are already dilated. Blood flow will therefore predominantly increase on the healthy side, leading to a marked asymmetry in rCBF.

Blood flow imaging with SPECT primarily allows the qualitative assessment of CBF. The assessment of the hemodynamic situation with that modality most often relies on the evaluation of the CBV/CBF ratio or the asymmetry index during baseline and induced hypercapnia. A common problem arises if there is bilateral or even more extended disease. In that case, the hypercapnia-induced blood flow increase may be uniformly impaired and not lead to marked asymmetries. In these situations, the CBV/ CBF ratio may be a more reliable measure for the hemodynamic situation. The most accurate evaluation is the quantitative assessment of the hemodynamics, as is possible with PET. A disadvantage in the clinical setting is that the accurate full quantification of cerebral blood flow requires the placement of an arterial catheter for the measurement of the tracer concentration in arterial blood. However, in a recent publication we showed that a simplified method may be sufficient for clinical purposes. The method is based on injected activity and does not require arterial blood sampling (8). Another important issue is the reproducibility of a method. We did an evaluation in 42 subjects who underwent quantitative perfusion scanning on two
consecutive days. On each day, three scans were performed 10 minutes apart. It was then assumed that the actual flow values during the same session were identical. So the difference between scan 2 – scan 1 and scan 3 – scan 2 in the same session is a measure of reproducibility. The results are shown in Figure 27.2B as a Bland-Altman plot (9) for gray and white matter. To determine gray matter flow, the flow values in all gray matter voxels were averaged. The range between the bottom and top horizontal line is defined by mean ± 2 SD. In gray matter, 2 SD was 17.2%. This means that a change in mean gray matter flow has to be larger than 17% in order to be considered significant with a 95% probability. In white matter, the corresponding value is 20.5%.

Examples of cerebrovascular evaluations are shown in Figures 27.3, 27.4 and 27.5. The benefit of surgical treatment of occlusive cerebrovascular disease is still inconclusive. One large international trial failed to demonstrate the effectiveness of EC-IC bypass surgery for preventing cerebral ischemia in patients with arteriosclerotic cerebrovascular disease (10). However, one criticism of that study is that the patients preoperative hemodynamic status was not
fully assessed. It is possible that a quantitative evaluation of the hemodynamic situation may better select those patients who can potentially benefit from bypass surgery. A more recent study involving 12 patients demonstrated hemodynamic improvement following bypass surgery using quantitative [¹⁵O]H₂O PET (11). A Japanese trial involving patients with low rCBF demonstrated no benefit from EC-IC bypass surgery in terms of stroke prevention, although a significant improvement of rCBF was found in the surgical group (12). It is obvious that larger studies are needed to resolve the issue.

SPECT allows the qualitative assessment of the hemodynamics using tracers like [^99mTc]HMPAO or [^99mTc]ECD. The major advantage of SPECT is its wide availability. Differences in the uptake pattern at baseline and during hypercapnia allow conclusions regarding the perfusion reserve. The cerebral blood volume can be assessed using ^99mTc-labeled erythrocytes. The disadvantages of SPECT compared with PET are as follows: