MRI markers in patients with ischaemic stroke : treatment and prognostic implications

Abstract

Stroke is currently the leading cause of disability in the world, with China having one of the highest stroke burdens. Cerebral small vessel disease (SVD) is the aetiological factor for 25% of strokes and for 45% of dementia cases. Due to the drawback of conventional imaging modalities, the small vessels cannot be visualised. Therefore, it is very difficult to observe pathological changes in the brain occurring as a result of SVD. These changes have been proposed as markers to allow prognostication in patients suffering from SVD. The newer imaging technologies like magnetic resonance imaging (MRI) help to diagnose these markers which include cerebral microbleeds (CMBs), perivascular spaces (PVSs), white matter hyperintensities (WMHs), and lacunes. Modalities like magnetic resonance angiogram (MRA) aid in the diagnosis and prognosis of large artery disease (LAD), which is a leading cause of ischaemic stroke (IS). LAD largely results from atherosclerosis, which is mainly caused by higher levels of low-density lipoprotein cholesterol (LDL-C). These levels can be easily monitored clinically and may give an insight to patient prognosis. These surrogate markers can aid treatment algorithms to increase long-term patient survival. Firstly, there are very few studies which have assessed the combined effect of total SVD score on adverse outcomes like recurrent stroke, recurrent IS, intracerebral haemorrhage (ICH), mortality, major adverse cardiovascular events (MACE) and poor modified Rankin scale (mRS). Secondly, with the advent of newer MRI sequences, various additional markers could be identified, which have not been a part of the total SVD score, and not explored in detail. Hence, I included brain atrophy, in addition to the current four markers in total SVD score. Thirdly, there are many risk factors associated with increased risk of SVD, with systolic blood pressure (SBP) being specifically highlighted. But the levels of post-event SBP levels that are associated with poor patient prognosis remain undefined. Fourthly, LDL-C is one of the most common measured parameters in post-event patient follow-up, and statin therapy is one of the primary treatment given during post-event. However, the optimal post-event LDL-C levels and potency of statin therapy in determining patient prognosis was not explored substantially. Therefore, for my thesis I aimed to determine the long-term prognostic implications of cerebral SVD in a Chinese cohort using a range of neuroimaging markers. I also aimed to explore the risk of vascular events in patients with IS and in those with severe cerebral SVD in relation to long-term blood pressure control. In addition, I assessed if low LDL-C is beneficial in IS patients without significant atherosclerosis. I have collated and analysed clinical and neuroimaging data from predominantly Chinese cohort comprising of 1003 IS patients recruited between 2008 to 2014 from the University of Hong Kong (HKU). All the patients had a cerebral MRI, with majority also receiving MRA at baseline. I interpreted a number of MRI markers of the cohort. Regular follow-up was done for all the patients for a mean of 5-6 years and the occurrence of adverse events including recurrent stroke, recurrent IS, ICH, MACE, mortality and poor mRS. Presence and burden of lacunes, CMBs, periventricular WMH (PWMH), subcortical/deep WMH (DWMH), basal ganglia (BG) and centrum semiovale (CS) PVSs, total SVD score and brain atrophy were determined for all patients. Post-event mean SBP levels and mean LDL-C levels were also recorded for the cohort. I observed several findings in this thesis which have clinical significance that would be beneficial to explore further. Firstly, on assessing the SVD markers individually, I found that a larger burden of CMBs and BG PVSs was associated with a greater risk of recurrent stroke, ICH, and MACE. Notably, I found that the long-term risk of ICH increased steeply with in patients with ≥5 CMBs. Although the absolute risks of recurrent ischaemic stroke in patients with ≥5 CMBs still outweighed that of ICH slightly, caution needs to be executed when initiating anti-thrombotic agents in patients with a high burden of CMBs. A higher burden of PWMH and brain atrophy was associated with poor MRS and mortality. The benefits of determining the total SVD score was also established as higher total SVD score was associated with increased risk of recurrent stroke, MACE, and ICH. However, I could not find any significant impact of lacunes and DWMH on risk of development of adverse outcomes. On assessment of patients who had a greater SVD burden, I found a U-shaped relationship between post-event mean SBP and adverse outcomes, with <120 and >140 mm Hg being prognostically associated with higher risk. Thus, the use of aggressive blood pressure lowering treatment is questioned, as the associated hypoperfusion may lead to more harm than benefit. On long-term follow-up, a mean LDL-C level of <1.80 mmol/L resulted in a lower risk of recurrent stroke and MACE. This relationship was present regardless of LAD status, IS subtypes, and age. This confers the need for using aggressive statin therapy to reduce the occurrence of adverse outcomes in patients.