Development of a novel MRI technique for imaging the ischaemic penumbra in experimental stroke

Robertson, Craig Alan (2011) Development of a novel MRI technique for imaging the ischaemic penumbra in experimental stroke. PhD thesis, University of Glasgow.

Abstract

In Scotland, stroke is the third most common cause of death behind heart disease and cancer. However, most strokes are not fatal and can cause severe disability, with one third of survivors still functionally dependent after one year. The advent of recombinant tissue plasminogen activator (rT-PA) as a thrombolytic modality revolutionised the treatment for ischaemic stroke, providing a treatment aimed to promptly restore nutritional blood flow to the ischaemic penumbra, a transient tissue state which is amenable to salvage. Crucially, patient ineligibility from a multitude of factors (including the narrow time window for benefit and the risk of intracranial haemorrhage) means that fewer than 10% of all stroke patients are thrombolysed.

Positive identification of penumbra is not employed in the current intravenous rT-PA administration strategy, which is instead based on two main prerequisites: stroke patients in whom intracerebral haemorrhage has been excluded with non-contrast computed tomography (CT) and who also present within 4.5 hours of symptom onset. The technical impracticalities and limited availability of the gold standard penumbral imaging modality, multitracer 15O positron emission tomography (PET), and the lack of standardised thresholds to identify penumbra using non-contrast CT have hindered the development and inclusion of routine brain imaging in the management of acute stroke patients. An alternative research tool which may potentially be used in clinical practice is magnetic resonance imaging (MRI) which defines penumbra on the basis of diffusion-perfusion (DWI/PWI) mismatch. However, this provides an imprecise measure of penumbra and fails to identify tissue viability.

Current PET-derived definitions of penumbra use metabolic indices such as oxygen extraction fraction (OEF) and the cerebral metabolic rate of oxygen (CMRO2), which are not fully incorporated into MR definitions. This thesis presents an alternative MRI method for identifying the metabolic penumbra in a rodent model of focal cerebral ischaemia. This utilises an MRI sequence similar to that used in functional MRI (fMRI) techniques, and uses 100% oxygen inhalation as a biotracer to detect penumbral tissue. Specifically, by using a blood oxygen level dependent (BOLD) T2*-weighted sequence in which changes in the deoxyhaemoglobin:oxyhaemoglobin ratio are detected - in conjunction with a transient hyperoxic challenge (Oxygen Challenge (OC) paradigm: 5 minutes breathing air followed by 5 minutes breathing 100% oxygen) - penumbral tissue can be distinguished from adjacent ischaemic core and benign oligaemia (Santosh et al, 2008).

Changes in CBF, cerebral blood volume (CBV), tissue oxygenation, and oxidative metabolism can all influence the T2* signal (Ramsay et al, 1993; Corfield et al, 2001), so it was important to evaluate the possibility that factors other than tissue metabolism were influencing the signal change during OC. An initial study was performed which showed that baseline CBF did not influence T2* signal response to OC, whilst a greater increase in the percentage change in arterial oxygen saturation following OC caused an increased magnitude in T2* percentage signal change in contralateral tissue and penumbra, but not in ischaemic core. Arterial oxygen levels (PaO2) affect the magnitude of the T2* signal change to OC, with lower baseline PaO2 levels amplifying the T2* signal response in metabolically active regions, implying that careful control of physiological variables may optimise the T2*OC technique.

The first validation study used [14C] 2-deoxyglucose autoradiography to determine the metabolic status of penumbra defined by T2*OC MRI. The results confirmed that glucose metabolism in the T2*OC-defined penumbra was comparable to contralateral values, whereas markedly different levels of glucose metabolism were evident in the ADC-derived ischaemic core and an adjacent region of increased 2DG phosphorylation. From this, it was concluded that metabolic information could be yielded from the ischaemic brain that may improve delineation of the penumbra using the OC technique.

As penumbral tissue must fulfil the fundamental criteria of being potentially salvageable and responsive to therapy, the consequences of reperfusion on the T2*OC-defined penumbra was tested. This study confirmed that T2*OC-defined penumbra displayed a T2* signal change significantly higher than contralateral tissue during ischaemia which subsequently returned to contralateral levels following reperfusion and did not progress to infarction when assessed at day 7 following stroke.

Finally, the spatiotemporal characteristics of the T2*OC-defined penumbra were investigated and compared with DWI/PWI mismatch-defined penumbra. Serial scanning demonstrated that T2*OC penumbra behaved in a similar manner to tissue defined by traditional mismatch criterion. The spatial location and tissue volumes of penumbra were similar with both methods, showing that, in animals where mismatch tissue volume reduced over time, T2*OC penumbra reduced similarly, and in animals where mismatch volume remained static over time, T2*OC-defined penumbra behaved similarly. Additionally, an interesting finding arose in the latter study which showed that ischaemic damage continues to progress beyond 4 hours following permanent MCAO, which may be relevant to the calculation of ADC and CBF thresholds used in defining DWI/PWI mismatch.

Collectively, the preclinical data support the potential of T2*OC to discriminate tissue compartments in acute stroke based on metabolic status which thereby provides an alternative and improved means of defining the ischaemic penumbra.

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Keywords: Stroke, MRI, Ischaemic penumbra, Oxygen Challenge
• Subjects: R Medicine > R Medicine (General); Q Science > QP Physiology
• Colleges/Schools: College of Medical Veterinary and Life Sciences > School of Psychology & Neuroscience
• Supervisor's Name: Macrae, Professor I Mhairi, McCabe, Dr. Christopher and Condon, Dr. Barrie
• Date of Award: 2011
• Depositing User: Mr Craig A Robertson
• Unique ID: glathesis:2011-3081
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 04 Jan 2012
• Last Modified: 10 Dec 2012 14:03
• URI: https://theses.gla.ac.uk/id/eprint/3081