Post-stroke dementia – a comprehensive review

Neuroimaging in PSD provides significant information about the anatomical substrate of the disorder and has an important role in PSD diagnosis. Further, it adds to prediction of cognitive decline after stroke; for example, hippocampal atrophy is a strong predictor for PSD outcome [12, 66]. Most acute stroke patients undergo CT brain imaging and, therefore, CT studies are representative of the whole clinical population. In everyday practice, CT is performed mainly to exclude intracerebral hemorrhage and certain stroke mimics (e.g., brain tumors), and can often show early signs of ischemia, as well as old stroke lesions. Additionally, the presence and extension of cerebral atrophy and white matter lesions can easily be seen on brain CT scans. Such neuroimaging characteristics may help predict subsequent cognitive impairment [67, 68].

MRI represents the most significant neuroimaging modality in PSD. If not contraindicated, MRI, rather than CT, is preferred for daily, routine clinical use as well as in research studies since it has higher specificity and sensitivity for detecting pathological substrates [66]. Neuroimaging standards with a generally accepted terminology allowing comparison of findings between different centers have been suggested (STandards for ReportIng Vascular changes on nEuroimaging, STRIVE) [69]. A large number of studies recognized MRI signs of cerebral SVD (lacunes, white matter hyperintensities, microbleeds, silent infarcts, white matter changes), as well as global cerebral atrophy and medial-temporal lobe atrophy as determinants and predictors of PSD. Vascular lesions associated with PSD are mainly found in subcortical brain areas, especially including sub-frontal and orbitofrontal white matter circuits. These lesions include single infarction in strategic areas such as the dominant thalamus or angular gyrus, deep areas of frontal lobe, and the left hemisphere, as well as brain infarcts in both hemispheres and volume-driven infarctions reaching a critical threshold of brain tissue loss or injury [70]. Additionally to size and location of the vascular lesion, the involvement of functional network fiber tracts, assessed by diffusion tensor imaging, may be crucial for cognitive impairment after stroke; its role is still being studied [71]. Although brain atrophy is a frequent finding in cerebral SVD, the pathophysiological mechanisms are not fully elucidated. In a recent study, Duering et al. [72] showed that cortical neurodegeneration following ischemia in the subcortical region probably represents pathophysiological mechanisms for cerebral atrophy in cerebrovascular disease; they showed that subcortical infarcts trigger focal thinning in connected cortical areas. Additionally, patients with mild VCI showed clear progressive gray matter atrophy in cortical (temporal and frontal) and subcortical (pons, caudate and cerebellum) regions after first-ever lacunar stroke, in contrast to patients without initial cognitive impairment [73].

Mild to moderate stroke patients with pre-existing white matter lesions are more vulnerable to cognitive impairment regardless of their new ischemic lesions [74]. Results from the SMART-MR study suggested that the interaction between brain atrophy and white matter hyperintensities or infarcts could aggravate cognitive decline [75]. There is evidence of the role of hippocampal mean diffusivity in the post-stroke cognitive state, above and beyond that of volume and connectivity of this structure [76].

Imaging of cerebral blood flow and cerebral glucose metabolism, by FDG-PET in particular, has been applied for several years in the differentiation between dementias [77]. However, functional and molecular imaging has a special role for research on the pathophysiology and the factors involved in development of PSD. Stroke and the deposition of amyloid in the cerebral cortex are both known risks for developing dementia [78, 79]. While animal models suggest that stroke-induced inflammation and amyloid deposition act synergetically [79], this relationship remains to be established in human stroke. Experimentally, large vessel infarcts or small striatal infarcts are larger in the presence of extracellular amyloid beta (A_?) deposits, which are fundamental markers of AD. Patients with minor cerebral strokes and moderate AD lesions will develop the clinical manifestations of dementia. A reasonable question is whether the stroke was just a signal and not necessarily a player in the pathogenesis of cognitive deterioration. Additionaly, small striatal infarcts in the presence of high levels of amyloid in the brain exhibit a progression in infarct size over time with an enhanced degree of cognitive impairment, AD-type pathology, and neuroinflammation compared with striatal infarcts or high amyloid levels alone. Recently, it has been stated that stroke is a specific vascular risk factor for cognitive impairment [80]. There is an overlap between VCI risk factors (hypertension, diabetes, and atherosclerosis) and those for stroke, as well as AD risk factors. Stroke may alter the clinical expression of a given load of AD pathology, but more studies are needed for definitive clarification of the relationships between stroke, AD, and cognitive detoriation.