The significance of fluoride uptake and inflammation within the atherosclerotic plaque: a PET/CT analysis

Background and aims: Atherosclerosis is a vessel disease, having a relatively early onset and a slow progression, and is currently the first cause for morbidity and mortality worldwide. Patients with atherosclerosis are usually classified according to their cardiovascular risk; however, tools to characterize the atherosclerotic plaque or to predict its progression are presently missing. In this thesis, we present a novel approach to this issue, by applying multimodal (CT- ad PET-based) imaging, in synergy with segmentation analysis, to the atherosclerotic plaques. Materials and Methods: Patients were recruited from three different databases retrospectively and then assigned to three different study populations. Population A consisted of 51 patients (19 females, mean age 699 years, range 49-82), submitted to 18F-NaF-PET/CT (NaF-PET). Each patient underwent at least 2 Naf-PET, spaced on average 14 months apart. In each patient, a VOI was placed on each visible CT plaque, using a CT-iso-contour approach; mean blood-pool normalized SUV (TBR), mean HU and Agaston-like calcification score (CS) were computed. TBR was compared with the mean percent variation of HU and CS, normalized for elapsed time between NaF-PETs (NDHU% and NDCS%, respectively). Whole-aorta TBR was then compared to whole-aorta NDCS%. Population B included 79 patients (51 women, 70.8 ±8 years) who underwent NaF-PET. Plaque analysis was performed as described above. An in-house software application was used to identify and segment the trabecular bone semi-automatically. TBR and HU of trabecular bone were compared to the ones of arterial plaques. Population C consisted of twenty-seven patients (12 males, mean age 69.4 ± 8, range 56-87), who underwent a 18F-FDG and a 68Ga-DOTATOC PET/CT within a two-weeks period Cardiovascular risk score was estimated in all patients; TBRmax and TBRmean was calculated in a large VOI, placed on the aorta, in FDG and DOTATOC scans. Results: In population A, mean HU and CS significantly increased from the first to the second PET/CT (p<0.001). A tight and direct correlation was noted between TBR in the plaques in the baseline PET and both NDHU% (R=0.67, p<0.01) and NDCS% (R=0,7, p<0.001). Whole-aorta TBR correlated with NDCS% in the entire vessel (R=0,85, p<0.001). In the population B, mean plaque density showed an inverse association with vertebral HU density (R=-0.56, p<0.01). Plaque and trabecular bone TBR were directly and closely correlated (R=0.63 and p<0.001). At univariate analysis, mean HU density of aortic plaque was not predicted by any of the cardiovascular risk factor or by age; conversely, it was related to its own TBR (p=<0.001) as well as by trabecular bone TBR. In population C, the mean of TBRmax was significantly higher in 68Ga-DOTATOC PET, when compared to FDG (5,7±3,1 Vs. 2±1,2, p<0.01). A tight and direct correlation was noted between FDG TBRmean and CV risk score (R=0.82, p<0.001), as well as between 68Ga-DOTATOC TBRmean and CV risk score (R=0.81, p<0.001). Average TBRmax of 68Ga-DOTATOC was slightly higher in DM patients when compared to the non-diabetic ones (6±2.1 Vs. 4,9±0,9, p<0.05). Conclusions: PET/CT with NaF can predict subsequent plaque evolution: in particular, plaque displaying a higher uptake have a greater progression of calcification at follow-up. Observing the behavior of skeletal bone might represent a new window for assessing the plaques´ characteristics. Inflammation within the plaque can be detected by 18F-FDG and by 68Ga-DOTATOC, the latter tracer might perform better, especially in diabetic patients. Overall PET techniques could display a great relevance in diverse research fields, such as assessment of therapy effectiveness and identification of vulnerable plaques; further study could allow the possibly enabling a patient-centered treatment and improving therapy outcomes as well as quality of life.application of this methods to larger population,