One of every 6 deaths in the USA in 2015 was caused by coronary artery disease (CAD). Traditionally, primarily anatomic considerations have been used to diagnose CAD. Fractional flow reserve (FFR), a physiological index of blood-flow reduction caused by coronary stenosis, has been shown by the FAME trials as a better predictor of clinical outcomes from coronary revascularization than that based on anatomy alone. PET-derived absolute myocardial blood flow (MBF), flow reserve (MFR) and relative flow reserve (RFR) have been shown to add clinical value in detecting CAD and risk assessment. Currently, PET measurements of MBF, MFR and RFR are not lesion specific, calculated either globally for the entire left ventricle (LV), or regionally to pre-defined vascular or segmental territories. This approach is limited by the intermixing of normal flow from normal regions with abnormal flow from abnormal regions thus reducing the measured degree of flow-impairment, diagnostic performance and culpable lesion location. We and others have shown that the variability alone of vessel pathway between patients leads to 18% misdiagnosis rate. We propose to develop algorithms to non-invasively measure MBF, MFR and RFR across specific coronary lesions for the entire coronary tree at least as accurately as those measured invasively during cardiac catheterization using fused coronary anatomy data obtained from CT coronary angiography (CTA) with dynamic PET (dPET) flow physiologic data. We hypothesize that our novel 3D fusion dPET/CTA approach will accurately and non- invasively predict lesion-specific severity as defined by invasive coronary angiography (ICA) FFR obtained with flow-wire/pressure-wire approaches. We anticipate that our dPET/CTA approach will be significantly more accurate than other existing non-invasive approaches. Exploiting our achievements in algorithm development, we will pursue our specific aims of 1) automating CTA myocardial border and vessel segmentation, 2) automating dPET/CTA 3D fusion to localize myocardial volumes of interest (VOIs) on dPET studies corresponding to the anatomical path of coronary vessels from CTA, and 3) calculating MBF and related flow parameters along coronary vessels using clinically accepted PET flow methods. Our dPET/CTA method will result in the following game-changing paradigm: 1) eliminate unnecessary ICAs in patients with no significant lesions, 2) avoid stenting physiologically insignificant lesions, 3) guide the PCI process to the location of significant lesions, 4) provide a flow-color-coded 3D roadmap of the entire coronary tree to guide bypass surgery, and 5) use less radiation and lower cost.
The aim of this work is to develop software tools to fuse coronary anatomy data obtained from CT coronary angiography with dynamic PET data (combination of anatomic and physiologic information) to noninvasively measure absolute myocardial blood flow, flow reserve and relative flow reserve across specific coronary lesions. These tools should reduce or eliminate unnecessary catheterizations and stenting and thus reduce patient risk, lower radiation exposure, and reduce the healthcare costs associated with unnecessary costly invasive treatments.
