Thesis

156 Part II Chapter 6 the interobserver reproducibility of the proposed method. All image analyses were performed in Osirix version 9.5 (Pixmeo SARL, Geneva, Switzerland). To assess whether a single Na18F uptake level could define pathological from benign bone uptake in our population, first the range of normal bone uptake was determined. For this purpose, a maximum of 5 spherical volumes of interest (VOIs) each with a 1-cm diameter (0.524 cm3, i.e. ~8 PET-voxels) were placed centrally in the vertebral bodies T12 and L5, mid sacrum, dorsally in the right iliac bone and intertrochanteric in the right femur, according to Rohren, et al., unless this bone was affected by any visual pathological condition [6]. In these latter cases, systematically alternative bone localizations were chosen (T11, L4, lower sacrum, left iliac bone and intertrochanteric in the left femur) if available [6]. Mean and standard deviation (SD) of the Na18F uptake concentration were determined (in Bq/mL) in each of these VOIs and normalized for decay-corrected injected Na18F-dose (in Bq), and corrected for either BW (in g), resulting in the standardized uptake value for BW (SUV) or Janmahasatian lean body mass (LBM, in g), resulting in the standardized uptake value for LBM (SUL) [6, 19]. The cut-off to discern normal from pathological Na18Fuptake in a single patient was defined as: − = ∑ ) . ( )5 + 3 ∙ . ( )5; ! ! "" #$ where ‘mean’ and ‘SD’ are the mean and SD of the VOI voxel values, respectively and n is the number of VOIs (maximum: 5). This defines pathological uptake per scan as the upper 0.13%, assuming a standard normal distribution of normal bone Na18F uptake voxel values. Preference for SUV or SUL was determined based on superior normalization to the VD of Na18F and defined as absence of correlation between cut-off SUV and BW and cut-off SUL and LBM, respectively. In case cut-off values in our cohort showed limited range, predefined as within 10% of its median value, a single cut-off would be accepted for all patients, otherwise a per scan cut-off would have been used [6, 7]. Quantification of FD burden on planar bone scintigraphy and Na[18F]F-PET/CT Baseline SBS was determined on planar bone scintigraphy (n=20) in supine position performed at start of the diagnosis of fibrous dysplasia, approximately 3 hours after injection of 550 MBq 99mTc-HDP (Siemens Symbia, matrix 256 x 1024, scan speed 24cm/min), and was blindly quantified by 2 experienced clinical readers (NMAD and MHR), differences were resolved in consensus [5]. The only intervention between baseline bone scan and Na[18F]F-PET/CT consisted of initiated bisphosphonate

RkJQdWJsaXNoZXIy MjY0ODMw