Altered Hemodynamic Responses in Patients After Subcortical Stroke Measured by Functional MRI

Abstract

Background and Purpose-Blood oxygenation level-dependent (BOLD) functional MRI (fMRI) is a promising method for defining brain recovery after stroke quantitatively. Applications thus far have assumed that the BOLD hemodynamic response in patients after stroke is identical to that in healthy controls. However, because of local vascular compromise or more diffuse vascular disease predisposing to infarction, this assumption may not be justified after stroke. We sought to test whether patients who have suffered a lacunar stroke show BOLD fMRI response characteristics identical to those of healthy controls. Methods-We measured the BOLD fMRI signal time course in the sensorimotor cortex contralateral to the affected hand with finger-or hand-tapping tasks for minimally or mildly impaired right-handed patients (nϭ12) after lacunar strokes causing limb weakness and for healthy controls (nϭ20). Results-With a right-handed sequential finger-tapping task, the rate of rise and maximum increase of the BOLD signal in the contralateral sensorimotor cortex were Ͼ30% lower (PՅ0.01) in the stroke patients. Similar relative decreases were found for the same task performed with the left hand. These changes were found in patients both in the hemisphere affected by stroke and in the unaffected hemisphere, suggesting that the BOLD fMRI time course differences observed arise from a diffuse functional pathology. The difference between patients and controls is not a result of age alone, since differences were not found between the younger (nϭ10; aged 22 to 38 years) and the older (nϭ10; aged 56 to 83 years) healthy controls. The effect also does not seem to be dependent on the specific hand movement task used. Conclusions-The magnitude of the BOLD fMRI response can be reduced in stroke patients even if infarcts do not involve the cortex. This may be a consequence of the stroke, but the observation that the BOLD signal time course is similar in the affected and unaffected hemispheres suggests that it also could result from preexisting pathophysiological changes in the cerebral microvasculature.

Visualization of Altered Neurovascular Coupling in Chronic Stroke Patients using Multimodal Functional MRI

“…However, patients without significant large-vessel disease (e.g., Patient 1) also showed impaired BOLD signal, and patients with persistent ICA occlusion (e.g., Patient 6) or previous ICA surgery due to known large-vessel disease (e.g., Patient 9) showed a positive BOLD signal. Previous studies have also shown impaired vascular reserve capacity in patients with suspected small-vessel disease due to hypertension (Ficzere et al, 1997), as well as lower BOLD response in patients with lacunar strokes (Pineiro et al, 2002) and/or white matter degeneration (Hund-Georgiadis et al, 2003), arguing in favor of an impairment of BOLD reactivity in both large-and small-vessel disease.…”

Reducing vascular variability of fMRI data across aging populations using a breathholding task

“…These include individual differences in the size and location of veins (Bandettini and Wong 1997;Cohen, et al 2004), extracranial arterial disease (Hamzei, et al 2003;Rother, et al 2002), small vessel cerebral disease (Pineiro, et al 2002), pulse or respiration differences (Dagli, et al 1999;, hematocrit concentrations (Levin, et al 2001), baseline CBF (Cohen, et al 2002). Several studies have also revealed that medications and food intake can affect the BOLD signal.…”

Spatio-temporal information analysis of event-related BOLD responses

“…However, application of this approach to other populations, such as aging and patient populations, has been a constant challenge [D'Esposito et al, 2003]. We expect that localization of task-related activation would be particularly sensitive to the fixed-HRF assumption, at higher association brain regions, and at subject populations, in which untypical HRFs are commonly observed [Pineiro et al, 2002;Rother et al, 2002;Hamzei et al, 2003]. Under theses circumstances, we predict that the information-theoretic method may be superior over the model based GLM analysis.…”