Is (18)F-FDG PET really a promising marker for clinically relevant atherosclerosis?

Bural et al (2013), retrospectively investigated 143 subjects who received whole body fluorine-18-fluorodeoxyglucose- positron emission tomography ((18)F-FDG-PET) imaging for the assessment of non-cardiovascular diseases. They reported an increase of (18)F-FDG-positive lesions in various aortic segments, which increased with age, and were more pronounced in subjects being aged below 50 years as compared to those above 50. Bural et al also found the highest segmental (18)F-FDG-uptake in the descending thoracic aorta, but not in the abdominal aorta, where the majority of the most severe atherosclerotic lesions essentially appear. In addition, they did not appreciate any significant gender difference. Despite the severe limitation that no correlation to vascular disease, risk factors, or any clinical parameter was available, this report again raises the question as to what positive (18)F-FDG imaging really reflects and whether it will ever reach the great expectations. Conventional radiotracers revealed an excellent experimental correlation, as well as morphology. Uptake ratios of symptomatic lesion vs. contralateral unaffected side were comparable between (111)In-platelets, (123)I-LDL and (18)FFDG. There was also a mass strategic correlation, but no individual prediction of events at all. Due to better statistics, image quality and solution PET imaging of atherosclerosis holds great promise. However, correlations between various tracers and vascular wall characteristics (and staining methodologies) in 1% cholesterol fed rabbits reveal that (18)F-FDG is not always the best tracer. Vascular foam cell content is reflected by (111)In-HIG > (125)I-oxLp(a) > (18)F-FDG > (125)I-LDL (Brammen L, Palumbo B, Lupattelli G et al. Unpublished data). A close correlation to Framingham risk score is for example not helpful, as this score has a low predictive value of only 0.6. The available clinical correlations between (18)F-FDG-uptake and arterial wall characteristics are poor. For example, Lederman RJ et al (2001) reported a correlation between (18)FFDG uptake with intima/media ratio, whereas no correlation was established in a paper by Ogawa M et al (2004). On the other hand, Laitinen I et al (2006) described a correlation between (18)F-FDG-uptake and calcifications, however, Tatsumi M et al (2003) did not observe this in his paper. The claim that inflammation and macrophage uptake of (18)F-FDG may be able to characterize and identify early atherosclerotic lesions has never been substantiated. Earlier studies reveal a negative correlation between (18)F-FDG uptake and smooth muscle cells, but a positive one with macrophages. The extent of uptake by different vascular wall cells (e.g. endothelial cells, smooth muscle cells, macrophages) in different atherosclerotic lesion types under various biochemical conditions has thus far not been extensively studied, neither in vitro nor in experimental or clinical work. Only one recent report does deal with this issue. Our preliminary studies show that the cellular uptake extremely varies depending on the local metabolic condition. For example, smooth muscle and endothelial cells, when exposed to pro-inflammatory cytokines, exhibit an extremely enhanced (18)F-FDG uptake while local hypoxia results in an opposite behavior. This is not observed in macrophages. Furthermore, when cultured cells were studied, uptake was severely dependent on the duration of incubation and the type of stimulation. This data indicates that (18)F-FDG uptake is enhanced in early foam cell formation, as well as in activated smooth muscle cells that eventually reach, under certain conditions, a comparable uptake. In addition, there is a lack of standardization and of prospective studies preventing reliable clinical interpretation. There seems to be only one consensus. There is no abnormal uptake of (18)F-FDG as well as of conventional tracers in the intact vascular wall and intra individual therapeutic intervention is truly reflected. The goal of non-invasive imaging in humans is to identify plaques at risk, an active lesion or the extent of the disease. As long as no prospective controlled data with other imaging modalities identifying vascular alterations defined per lesion and not per segment are available, it seems very unlikely that (18)F-FDG may significantly succeed in this particular indication.