Abstract
1075
Objectives: To assess the utility of quantitative amyloid plaque deposition using SUV ratio (SUVr) analysis in corroborating and enhancing visually interpreted 18F-florbetaben (FBB) PET/CT scans.
Methods: From a large cohort of patients who underwent FBB PET/CT imaging as part of an IRB approved research study at our institution, and whose scans were visually interpreted and classified based on the brain amyloid plaque load (BAPL) scoring scheme, all studies visually interpreted to date as BAPL3 (n=16), BAPL2 (n=17) as well as an equivalent number of randomly chosen studies interpreted as BAPL1 (n=17) were quantitatively analyzed. Imaging was performed with Philips Gemini PET/CT scanners (TF-16 and TF-64) following the injection of 300 MBq of 18F-florbetaben and a 70-minute incubation time. Unenhanced CT images were acquired from the top of the head to the upper neck followed by PET images, which were reconstructed in the axial, sagittal and coronal views and visually interpreted by an experienced dual-boarded Nuclear Medicine / Radiology Physician. Subsequently, the scans were quantitatively analyzed for amyloid plaque deposition using a commercial software package (syngo.via MI; Siemens Healthineers). Regional (anterior cingulate gyrus, frontal lobe, occipital lobe, parietal lobe, posterior cingulate gyrus, temporal lobe) as well as composite cortical region SUVr values were generated using the whole cerebellum (CER) as the normalization region. Quantitative results were analyzed with ANOVA and correlation plots to evaluate for possible outliers in the three groups.
Results: The mean composite SUVr values for the visually classified cases were 1.32, 1.44 and 1.97 for BAPL1, BAPL2 and BAPL3, respectively. All BAPL ratings were statistically different in terms of their composite SUVr values. Further, these findings were consistent for all the cortical regions of interest used for the comparison while maintaining whole cerebellum (CER) as the normalization region.
Conclusions: Visual interpretation of FBB cases into BAPL1, BAPL2 and BAPL3 was corroborated and enhanced with SUVr quantitation of amyloid plaque regional and whole brain deposition using commercial software. The quantitative analysis served to solidify the visual assessment of the BAPL3 cases, and in some cases to enable the distinction between BAPL1 and BAPL2 cases that was otherwise difficult with visual inspection. Quantitation is a useful adjunct tool to visual interpretation of FBB PET/CT imaging and together with clinical correlation can aid in a more comprehensive evaluation of patients with cognitive impairment.