Persistent Steal Physiology during a Ramp Cerebrovascular Reactivity Assessment Correlates with Severe Cortical...

Steal physiology, a marker of severe cerebrovascular impairment, is associated with cortical thinning, a morphologic phenomenon seen in patients with chronic supply artery steno-occlusive disease. This link has been previously established only with a sudden "step" hypercapnic stimulus. The significance of steal patterns derived from a gradual "ramp" stimulus, and their association with structural brain injury, is unclear. We investigated how steal patterns derived from both step and ramp stimuli compare and relate to the severity of cortical thinning in patients with steno-occlusive disease.

Thirty-two patients with unilateral cortical steal, characterized by an asymmetric paradoxical reduction in blood flow during vasodilatory stimuli due to impaired vascular reserve in one hemisphere, were retrospectively identified from a prospectively maintained database. Cerebrovascular reactivity (CVR) was mapped using BOLD MRI during sequential step and ramp CO₂ stimuli from a sequential gas delivery system. The spatial extent of cortical steal (volume with negative CVR) was quantified and correlated with interhemispheric cortical thickness differences for each stimulus type.

The spatial extent of cortical steal physiology derived from the ramp stimulus was consistently smaller, averaging 81% of that seen with the step stimulus (P = .004). Both stimulus approaches demonstrated significant correlations with cortical thinning (step: r = 0.56, P < .001; ramp: r = 0.49, P = .004). Patients exhibiting steal during both stimulus types showed more pronounced cortical thinning compared with those demonstrating steal only during step stimuli, suggesting that persistent steal across both protocols indicates more severe cerebrovascular impairment.

The persistence of steal physiology seen during a gradual ramp stimulus reveals a more serious hemodynamic impairment than steal observed only during a sudden step challenge in patients with steno-occlusive disease. While the step stimulus is highly sensitive, its greater spatial extent of steal includes all hemodynamically vulnerable tissues, even those with slower but potentially effective collateral responses. The ramp stimulus gives these tissues more time for vascular response to evolve fully, thereby isolating persistent steal that represents complete failure of collateral circulation. This failed adaptation is associated with more severe structural brain injury. Therefore, multistimulus CVR protocols may offer a more nuanced approach to risk stratification.