Abstract
Introduction: Sickle Cell Disease (SCD) is frequently complicated by stroke. Although transcranial Doppler effectively identifies children at risk for stroke, adult patients do not benefit from this test. In SCD, chronic hemolytic anemia leads to cerebral vasodilation, elevated cerebral blood flow (CBF) and subsequently, impaired cerebrovascular reserve (CVR) capacity. CVR represents the maximum increase in CBF in response to metabolic stressors. CVR mapping is a promising imaging biomarker for stroke risk assessment, potentially identifying patients with preclinical hemodynamic impairment. Our laboratory performs CVR mapping by measuring CBF prior to and following maximum cerebral vasodilation with acetazolamide (ACZ). The primary aim of this study was to quantitatively assess CVR in adult patients with SCD compared to healthy controls.
Methods: Adult SCD patients (HbSS/HbSβ0) were recruited for this IRB-approved study with ACZ-induced vasodilation and venous blood sampling. Patients with a history of clinically overt stroke were excluded. Controls were selected from patients' friends and family members without SCD and matched on age, sex, and race. The following MRI images were acquired at 3T (Philips Healthcare, Best, NL): time of flight MRA for visualization of major cerebral vessels, T2 FLAIR for manual segmentation of white matter hyperintensities (WMHs), and pseudo-continuous arterial spin labeling (ASL) for CBF assessment. CBF was measured at baseline and 10min post ACZ (16mg/kg intravenous infusion over 3min). MRI images were processed with the ExploreASL toolbox, to obtain registered maps of quantified CBF and CVR (% change in CBF). We measured T1blood in each subject to improve quantification accuracy as these values can differ in SCD. Group comparisons were performed using non-parametric two-sample tests. Correlations were characterized by Spearman's rho (ρ). P<0.05 was considered significant. Median values with interquartile range (IQR) are reported for non-normally distributed variables, otherwise mean and standard deviation are reported.
Results: 30 patients with SCD (mean age 33±12y,19M) and 11 controls (mean age 37±15y, 6M) were included in this cross-sectional controlled cohort study. Mean hemoglobin levels in patients with SCD were 8.8±1.4 g/dL and in healthy controls were 13.7±1.3 g/dL. Patients with SCD had higher baseline CBF compared to healthy controls (median 73(IQR:25) vs 42(IQR:6) mL/100g/min, p<0.001, F1a). CBF was inversely related to hemoglobin (ρ=-0.84, p<0.001, F1b). ACZ elicited an increase in CBF(p<0.001) which was similar in magnitude in both groups (patients 29±16, controls 35±11, mL/100g/min, F1c), resulting in a lower mean CVR in patients compared to controls (41±24% vs 81±27%, p<0.001, F1d). Baseline CBF predicted CVR (ρ=-0.68, p<0.001, F1e). WMHs were present in both groups, and WMH volume correlated with age (ρ = 0.54, p<0.001). There were no statistical associations between WMH volume and CVR (ρ = -0.15, p=0.4) or CBF(ρ = 0.02, p=0.9), however, two patients with the largest WMH volume (>6 mL) had the lowest CVR (<20%).
Discussion: This study shows that regional CVR can be measured using ASL-MRI with ACZ challenge. ACZ was well-tolerated and elicited a robust cerebrovasodilatory response in both groups. However, the relative increase in oxygen delivery (CVR) was much lower in SCD patients, which suggests that patients with SCD have nearly maximal cerebrovascular dilation at baseline. Our data suggest that chronic vasodilation due to anaemia has let to outward remodelling of vessels in adult patients, permitting a larger vascular bed to compensate their anemia. Our premise is that resting O2 delivery is normal in SCD patients; we showed that increasing CBF overcomes severe anemia. However, high resting CBF limits the brain's ability to recruit additional O2 under times of stress. SCD patients have many transient interruptions in O2 delivery including aplastic crisis, splenic sequestration and sleep apnea, as well as metabolic stressors such as fever, sickle cell crisis, infection, and seizure. While baseline CBF predicted global CVR, ASL provides information regarding regional O2 delivery that may offer insight into distribution of ischemic white matter damage. Further study is needed to determine the impact of blood transfusions and hydroxyurea on CVR and whether there is a critical CVR threshold that predicts stroke risk.
Disclosures
No relevant conflicts of interest to declare.
Regional oxygen extraction predicts border zone vulnerability to stroke in sickle cell disease
