Relevance of autoregulation in the diagnosis of coronary artery disease with CT-FFR

Vasodilation and vasoconstriction of small resistance vessels in the coronary microcirculation are key mechanisms for regulating coronary blood flow. This dynamic regulation acts as a reserve system that ensures adequate perfusion under varying physiological conditions and metabolic demands. In the presence of coronary artery stenosis, part of this vasodilatory reserve may already be used, triggered by pressure-flow autoregulation mechanisms that maintain normal blood flow at rest. Fractional flow reserve (FFR), measured during pharmacologically induced maximum vasodilation/hyperemia (typically with adenosine), remains the gold standard for diagnosing coronary artery disease. Computed tomography-derived FFR (CT-FFR) provides noninvasive FFR estimates and is increasingly recommended as a gatekeeper to invasive evaluation. Most current CT-FFR models assume a fixed microvascular response to adenosine, treating vasodilation capacity as constant and optimal for all patients and lesions. Using intracoronary pressure measurements, we quantified how stenosis severity affects the microcirculatory response to adenosine. Based on these findings, we developed a CT-FFR algorithm that incorporates autoregulation. In comparison across 280 lesions, the autoregulation-based model reduced prediction variability (SD ¼ 0.1006 vs. 0.1167, P value ¼ 0.013) and improved the coefficient of determination (R2 ¼ 0.398 vs. 0.179) relative to a fixed-response model, using invasive FFR as reference. However, the area under the curve (AUC) remained unchanged. Sensitivity analyses showed that autoregulation effects were minor in lesions with CT-FFR > 0.7. Feature analysis showed that vessel location, patient sex, and heart rate are associated with the microvascular response to adenosine, emphasizing the need to consider these factors in the combined assessment of epicardial and microvascular disease.