Elevations in MR Measurements of Whole Brain and Regional Cerebral Blood Flow and Oxygen Extraction Fraction Suggest Cerebral Metabolic Stress in Children with Sickle Cell Disease Unaffected By Overt Stroke

Positron emission tomography (PET) studies have shown that the brain increases cerebral blood flow (CBF) and oxygen extraction fraction (OEF, the fraction of oxygen brain tissue extracts from blood) when oxygen delivery is compromised in adult ischemic stroke. Children with sickle cell disease (SCD) have higher CBF compared to healthy children, suggesting that autoregulatory mechanisms, compensating for compromised oxygen delivery, may underlie the pathophysiology of ischemic stroke in SCD.  Until now, evaluation of cerebral oxygen metabolism in children with SCD has been limited to measurement of CBF due to the radiation risks of PET. We used a MR sequence that measures voxel-wise OEF to test our hypothesis that children with sickle cell disease will have elevated whole brain and regional OEF when compared to typically developing, sibling controls without SCD.

Thirty-six participants, 8 controls and 28 with SCD (26 HbSS and 2 HbSB⁰), ages 5-21 years, were recruited from St. Louis Children's Hospital.  Participants underwent brain MRI with measurement of CBF via pseudo-continuous arterial spin labeling and OEF via a novel processing of asymmetric spin echo sequence that measures tissue deoxyhemoglobin. CBF and OEF maps were individually co-registered to corresponding T1 images with FMRIB's Linear Image Registration Tool, and gray and white matter were segmented with FMRIB's Automated Segmentation Tool. Visual inspection identified a region of high OEF within the deep white matter of the frontal and parietal lobes in the majority of subjects (figure 1 a,b).  OEF maps from control and SCD participants were coregistered and averaged into a single map, and then an OEF threshold of 47.5% was applied to demarcate this "high OEF region" (figure 1c). Hemoglobin (Hb) and hematocrit were obtained in SCD participants, while these values were assumed for the controls. Arterial oxygen content (CaO₂) was calculated as 1.36 x Hb x SpO₂. Comparisons were made with a Mann-Whitney U test or Student's t-test. Bivariate correlations were tested with Kendall's tau b. Bonferroni correction was used in determining significance. Multivariate linear regression modeling with block entry described covariates associated with CBF.

The control and SCD cohorts did not differ in age, gender or SpO₂. SCD participants demonstrated higher whole brain, gray matter and white matter CBF and OEF when compared to controls (table 1, figure 1a-b), but there was no difference in whole brain or segmented measures of CBF and OEF between SCD participants with structurally normal brain MRIs (n=16) and silent infarcts (n=12). SCD participants' OEF was higher within the "high OEF region" when compared to controls (table 1), but the regional OEF did not differ between SCD participants with structurally normal brain MRIs versus silent infarcts. Whole brain CBF negatively correlated with age (b = -0.554, p < 0.001), while whole brain OEF did not (b= 0.014, p = 0.921). Lower CaO₂ correlated with higher whole brain CBF (b = -0.329, p < 0.016) and higher whole brain OEF (b = -0.587, p < 0.001).  CaO₂ remained a predictor (β = -0.38, p = 0.009) of CBF when controlling for age (β = -0.63, p < 0.001).

We report the first whole brain, segmented and regional analysis of oxygen metabolism, including CBF and OEF, in a pediatric SCD cohort unaffected by overt stroke.  Elevation in both CBF and OEF in response to low arterial oxygen content suggests children with SCD are chronically compensating to meet the brain's metabolic demands. OEF is more elevated in children with SCD when compared to healthy controls within the "high OEF region," which coincides with previously reported locations of silent infarcts. We propose that the "high OEF region" provides a tissue signature of vulnerable, metabolically stressed brain that is at risk for future stroke.