The longitudinal changes of BOLD response and cerebral hemodynamics from acute to subacute stroke. A fMRI and TCD study

Abstract

Background: By mapping the dynamics of brain reorganization, functional magnetic resonance imaging MRI (fMRI) has allowed for significant progress in understanding cerebral plasticity phenomena after a stroke. However, cerebro-vascular diseases can affect blood oxygen level dependent (BOLD) signal. Cerebral autoregulation is a primary function of cerebral hemodynamics, which allows to maintain a relatively constant blood flow despite changes in arterial blood pressure and perfusion pressure. Cerebral autoregulation is reported to become less effective in the early phases post-stroke. This study investigated whether any impairment of cerebral hemodynamics that occurs during the acute and the subacute phases of ischemic stroke is related to changes in BOLD response. We enrolled six aphasic patients affected by acute stroke. All patients underwent a Transcranial Doppler to assess cerebral autoregulation (Mx index) and fMRI to evaluate the amplitude and the peak latency (time to peak-TTP) of BOLD response in the acute (i.e., within four days of stroke occurrence) and the subacute (i.e., between five and twelve days after stroke onset) stroke phases.

Results: As patients advanced from the acute to subacute stroke phase, the affected hemisphere presented a BOLD TTP increase (p = 0.04) and a deterioration of cerebral autoregulation (Mx index increase, p = 0.046). A similar but not significant trend was observed also in the unaffected hemisphere. When the two hemispheres were grouped together, BOLD TTP delay was significantly related to worsening cerebral autoregulation (Mx index increase) (Spearman's rho = 0.734; p = 0.01).

Conclusions: The hemodynamic response function subtending BOLD signal may present a delay in peak latency that arises as patients advance from the acute to the subacute stroke phase. This delay is related to the deterioration of cerebral hemodynamics. These findings suggest that remodeling the fMRI hemodynamic response function in the different phases of stroke may optimize the detection of BOLD signal changes.

Figure 1

Ischemic Lesions. Axial Diffusion Weighted MR images of enrolled patients. The left side of the figure corresponds to the left side of the brain.

Figure 2

BOLD signal in controls and patient 2 performing language paradigm 1. (A) HRF extracted from the peak voxel in the bilateral auditory cortex of six healthy control subjects performing language paradigm 1. Plots represent the mean contrast estimate across subjects (y-axis) within each time bin (x-axis). The dark grey bar indicates the time bin with the highest contrast estimate as an estimation of the TTP latency. (B) HRF extracted from the peak voxel in the bilateral auditory cortex of patient 2 at day 2 (upper row) and day 8 (lower row). Plots represent mean contrast estimate across stimuli (x-axis) within each time bin (y-axis). From day 1 to day 8, TTP latency increased (from time bin 5 to time bin 6) and amplitude decreased slightly in both hemispheres.

Figure 3

BOLD signal in controls and patient 6 performing language paradigm 2. (A) HRF extracted from the peak voxel in the bilateral auditory cortex of six healthy control subjects performing language paradigm 2. Plots represent the mean contrast estimate across subjects (y-axis) within each time bin (y-axis). The dark grey bar indicates the time bin with the highest contrast estimate as an estimation of the TTP latency. (B) HRF extracted from the peak voxel in the bilateral auditory cortex of patient 6 at day 1 (upper row) and day 9 (lower row). Plots represent mean contrast estimate across stimuli (x-axis) within each time bin (y-axis). From day 1 to day 9, TTP latency increased in both hemispheres (from time bin 6 to time bin 7) and amplitude decreased in the lesioned hemisphere.