Complications of hemorrhagic and ischemic stroke : a CT perfusion evaluation

Abstract

In this thesis the use of CT-perfusion (CTP) imaging in the evaluation of the most severe complications of subarachnoid hemorrhage (SAH)) and ischemic stroke was explored. These complications are delayed cerebral ischemia (DCI) after SAH and damage to the blood-brain barrier (BBB) after ischemic stroke causing hemorrhagic transformation (HT) and malignant edema (ME). In part 1 of this thesis we evaluated the pathogenesis of DCI with CTP, the diagnosis of DCI with CTP, and DCI treatment evaluation with cerebral perfusion measurements. In part 2 we looked at the optimal use of CTP for BBB permeability (BBBP) measurement, factors influencing BBB integrity in stroke and non-stroke patients, BBB integrity in infarcted, ischemic and non-ischemic tissue, and the predictive value of BBBP measurements for HT and ME. Part 1: With CTP cerebral perfusion can be accurately estimated in patients without intracranial pathology. There are normally small differences between perfusion in different cerebral flow territories. Deviations from these normal differences may indicate pathology. In SAH patients vasospasm causes a decrease in perfusion in the flow territory behind the spastic segment. However, large vessel vasospasm is not present in almost half of DCI patients. This suggest that vasospasm alone is not sufficient to cause DCI and that a multifactorial origin of DCI is more plausible. Over time absolute perfusion parameters gradually worsen towards DCI while perfusion asymmetry worsens more prominently and occurs after absolute values have already worsened. This suggests that SAH patients with diffusely worse absolute perfusion parameters may be more susceptibly to additional changes in perfusion. Qualitative interpretation of CTP colormaps makes it possible to distinguish patients with DCI from patients with other causes of deterioration and quantitative perfusion measurements can distinguish patients with DCI from clinically stable patients. Currently there is no good evidence that CBF improves due to treatment with Triple-H or its separate components. Of all components, induced hypertension seems to be the most promising. Part 2: BBB damage is the mechanism by which HT and ME occur after acute ischemic stroke. To calculate BBBP from CTP values the Patlak model can be applied to the delayed phase of the CTP acquisition (and not the first-pass). A second bolus can be used to increase anatomical coverage of BBBP measurements with CTP. In patients without ischemic stroke BBBP is increased in the presence of hypertension and diabetes type II. Aspirin use seems to protect against BBB damage. In patients with ischemic stroke BBBP is increased in infarcted and ischemic brain tissue compared to non-ischemic tissue on the admission CTP scan. Patients with more than 50% collateral filling have lower BBBP values than patients with less collateral filling. Good collateral supply therefore seems to protect against BBB damage. Patients in whom severe BBBP damage leads to HT or ME can be accurately identified with BBBP measurements from the delayed phase of CTP on admission. Only patients with abnormally elevated BBBP, older than 65 years that received rtPA developed HT or ME in our population.