Qualitative and quantitative analysis of 18F-GP1 positron emission tomography in thrombotic cardiovascular disease

Abstract

Abstract 18F-GP1 is a novel highly specific radiotracer that binds to activated platelets and thrombus. We aimed to establish the observer repeatability of coronary, carotid and cerebral 18F-GP1 uptake in patients presenting with acute myocardial infarction or ischaemic stroke. Forty-three patients presenting with acute myocardial infarction or ischaemic stroke underwent hybrid positron emission tomography (PET) and computed tomography (CT) angiography. Qualitative and quantitative assessment of 18F-GP1 uptake was performed on coronary arteries, carotid arteries and brain parenchyma. Qualitative uptake of 18F-GP1 had excellent intraobserver and interobserver agreement, with complete agreement for the presence or absence of visual 18F-GP1 uptake. For quantitative analysis, there were excellent intraclass correlation coefficients for intraobserver repeatability for coronary artery, carotid artery and brain parenchymal SUVmax and TBRmax measurements (all ≥ 0.92). Coronary artery and brain parenchymal analyses showed the strongest agreement in SUVmax values with mean biases of − 0.04 (limits of agreement − 0.21 to 0.20) and 0.02 (limits of agreement − 0.29 to 0.32) respectively. There was good interclass correlation coefficients for interobserver repeatability for coronary artery, carotid artery and brain parenchymal SUVmax and TBRmax measurements (all ≥ 0.89). The strongest interobserver agreement was seen with brain parenchymal SUVmax (mean SUVmax 1.95 ± 0.94) and TBRmax (mean TBRmax 9.55 ± 6.56) with mean biases of − 0.05 (limits of agreement − 0.37 to 0.27) and 0.04 (limits of agreement − 0.59 to 0.52) respectively. Visual qualitative and quantitative 18F-GP1 PET-CT image analyses provide robust and repeatable measurements of activated platelets and thrombi within the coronary arteries, carotid arteries and brain parenchyma.