Effect of Blood Transfusion on Cerebral Hemodynamics and Vascular Topology Described by Computational Fluid Dynamics in Sickle Cell Disease Patients.

The main objective of this study was to demonstrate that computational fluid dynamics (CFD) modeling can be used to study the contribution of covert and overt vascular architecture to the risk for cerebrovascular disease in sickle cell disease (SCD) and to determine the mechanisms of response to therapy such as chronic red blood cell (cRBC) transfusions. We analyzed baseline (screening), pre-randomization and study exit magnetic resonance angiogram (MRA) images from 10 (5 each from the transfusion and observation arms) pediatric sickle SCD participants in the silent cerebral infarct transfusion (SIT) trial using CFD modeling. We reconstructed the intracranial portion of the internal carotid artery and branches and extracted the geometry using 3D Slicer. We cut specific portions of the large intracranial artery to include segments of the internal carotid, middle, anterior, and posterior cerebral arteries such that the vessel segment analyzed extended from the intracranial beginning of the internal carotid artery up to immediately after (~0.25 inches) the middle cerebral artery branching point. Cut models were imported into Ansys 2021R2/2022R1 and laminar and time-dependent flow simulation was performed. Change in time averaged mean velocity, wall shear stress, and vessel tortuosity were compared between the observation and cRBC arms. We did not observe a correlation between time averaged mean velocity (TAMV) and mean transcranial Doppler (TCD) velocity at study entry. There was also no difference in change in time average mean velocity, wall shear stress (WSS), and vessel tortuosity between the observation and cRBC transfusion arms. WSS and TAMV were abnormal for 2 (developed TIA) out of the 3 participants (one participant had silent cerebral infarctions) that developed neurovascular outcomes. CFD approaches allow for the evaluation of vascular topology and hemodynamics in SCD using MRA images. In this proof of principle study, we show that CFD could be a useful tool and we intend to carry out future studies with a larger sample to enable more robust conclusions.