Developmental patterns of CBF and BOLD responses to visual stimulus

To investigate the developmental changes of cerebral blood flow (CBF) and hemodynamic responses to changing neural activity, we used the arterial spin label (ASL) technique to measure resting CBF and simultaneous CBF / blood-oxygen-level dependent (BOLD) signal changes during visual stimulation in 97 typically developing children and young adults (age 13.35 [6.02, 25.25] (median [min, max]) years old at the first time point). The longitudinal study protocol included three MRIs (2.7 ± 0.06 obtained), one year apart, for each participant. Mixed-effect linear and non-linear statistical models were used to analyze age effects on CBF and BOLD signals. Resting CBF decreased exponentially with age ( p = 0.0001) throughout the brain, and developmental trajectories differed across brain lobes. The absolute CBF increase in visual cortex during stimulation was constant over the age range, but the fractional CBF change increased with age ( p = 0.0001) and the fractional BOLD signal increased with age ( p = 0.0001) correspondingly. These findings suggest that the apparent neural hemodynamic coupling in visual cortex does not change after age six years, but age-related BOLD signal changes continue through adolescence primarily due to the changes with age in resting CBF.

Abstract W MP114: Arterial Spin Label Perfusion Imaging in Acute Neonatal Stroke Reveals Hyperperfusion in Association With Cerebral Ischemic Injury

Introduction: Perfusion magnetic resonance imaging (pMRI) in adults reveals a hypoperfused region, the ischemic penumbra, related to the fixed ischemic injury shown by diffusion weighted imaging (DWI) in acute stroke. Neonatal stroke has pathophysiology and optimum treatment that remain elusive. We hypothesized that altered regional perfusion accompanies the ischemic injury in focal neonatal stroke. Methods: Neonates (0-28 days of life) underwent MR imaging at 3T fieldstrength. pMRI employed a pseudocontinuous arterial spin labeling (PCASL) sequence with multi-slice echo planar readouts at 3x3x5mm3 resolution with TE/TR = 12ms/3.5s, labeling time = 1.6s, a post-labeling delay = 1.5s, 9 axial slices acquired, scan time ~5 min. Forty label/control image pairs were acquired, subtracted and averaged to obtain contrast maps proportional to quantified cerebral blood flow (CBF). DWI, susceptibility weighted imaging and T2-weighted series were obtained with standard protocols. Acute stroke was classified as arterial or venous. Core infarction was considered to lie in hypointense regions on DWI apparent diffusion coefficient maps. Regional perfusion signal was compared to the ischemic core on DWI and to the homologous uninvolved region in the contralesional hemisphere. Results: Data were collected on 20 neonates, (13 males; median/mean gestational age=38/38.5 weeks) with acute stroke (11 arterial ischemic stroke (AIS); 9 venous) scanned at median postnatal age=1.5 days. Evidence of regional hyperperfusion existed in 10/11 (91%) neonates with AIS and 4/9 (44%) with venous stroke. Hyperperfusion was part of a heterogeneous pattern of hypo- and hyperperfusion in 4 infants with AIS and 3 infants with venous infarct. Electroencephalography data obtained at presentation were available for 19 (95%) of the neonates. Fourteen infants (70%) had either electrographic seizure or focal sharp waves in the same hemisphere as the acute infarction. Conclusion: Arterial spin labeling pMRI can be successfully obtained in acute neonatal stroke. Unlike adults, pMRI often reveals hyperperfusion rather than hypoperfusion in neonates. This hyperperfusion may be due to early spontaneous reperfusion or to the neuronal hyperexcitability of stroke-associated seizure.