Cerebrovascular reactivity (CVR) is the change in cerebral blood flow (CBF) in response to a change in a vasoactive stimulus. Paradoxical reductions in CBF in response to vasodilatory stimulation ('steal') are associated with vascular pathology. However, a pathophysiological interpretation of 'steal' requires a comprehensive conceptual model linking pathology and changes in blood flow. Herein, we extend a simple model explaining steal published in the late 1960s by incorporating concepts of CBF regulation from more recent studies to generate a comprehensive dynamic model. The main elements of the model are: (a) the relationship between changes in CBF and the arterial partial pressure of carbon dioxide (PaCO₂) in healthy vascular regions is sigmoidal; (b) vascular regions vasodilate to compensate for decreased perfusion pressure, leading to (c) an encroachment on vasodilatory reserve and, reduced CVR; (d) a vasodilatory stimulus may increase CBF capacity above the flow capacity of major cerebral blood vessels; and (e) this limitation induces competitive intra-cerebral redistribution of flow from territories with low vasodilatory reserve to those with high reserve. We used CVR measurements generated by applying precise, computer-controlled changes in PaCO₂ as the vasoactive stimulus, and measured blood oxygen level dependent (BOLD) MRI signals as high resolution surrogates of CBF to test predictions derived from this model. Subjects were 16 healthy adults and 16 patients with known cerebral steno-occlusive diseases. We observed regional sigmoidal PaCO₂-BOLD response curves with a range of slopes; graded changes in PaCO₂ resulted in redistributions of BOLD signal consistent with the known underlying vascular pathology and predictions of the model. We conclude that this model can be applied to provide a hemodynamic interpretation to BOLD signal changes in response to hypercapnia, and thereby aid in relating CVR maps to pathophysiological conditions.
Keywords: Carbon dioxide; Cerebrovascular reactivity; Humans; Reverse steal; Steal.
Copyright © 2014 Elsevier Inc. All rights reserved.