Background PET is a nuclear imaging technique that depicts functional processes within the body based on gamma rays. The biologically active molecule used for PET is the bone-seeking agent 18F labeled Fluoride (18F-F). The concentrations of tracer reflect tissue metabolic activity of regional bone perfusion and bone remodeling. PET scans are increasingly read alongside CT or MRI scans, with the combination (co-registration) giving both anatomic and metabolic information on the nature of the structure and it biochemical activities.
Objectives To test the performance of integrated 18F-Fluoride positron emission tomography, magnetic resonance imaging (PET/MRI) and computer tomography (CT) of the whole spine and to compare the inflammatory, structural and metabolic findings in patients with ankylosing spondylitis (AS).
Methods 13 AS patients (6 male, 7 female, mean age 37.8±11.4 years, all BASDAI>4, no anti-TNF treatment) underwent a 3-Tesla MRI and integrated PET/MRI 40 minutes after injection of a mean dose 157 MBq of 18F-Fluoride of their whole spine. Low-dose CT examinations were available from 5 of those patients. Two readers scored all images independently and the lesions where both readers showed agreement were considered for analysis. Inflammatory activity (bone marrow edema, BME), structural lesions (fat deposition), focal 18F-Fluoride uptake and new bone formation detected on CT were recorded on the level of a vertebral quadrant (VQ), where one vertebral body consisted by 4 VQs (superior anterior and posterior and inferior anterior and posterior).
Results Acquisition of whole-spine 18F-Fluoride PET/MRI was successful in all patients. The readings of both readers agreed in 96% of all lesions detected, showing an excellent performance of the applied techniques. A total of 1,196 VQs could be analysed by MRI and 460 VQs could be additionally analyzed by CT. A total of 118/1,196 VQs (9.9%) showed BME and 218/1,196 (18.2%) showed fat deposition on 3T-MRI, while 65/1,196 VQs (5.4%) showed focal 18F-Fluoride uptake in PET/MRI and 68/460 (14.8%) showed syndesmophytes on CT. Association to quadrants increased PET/MRI uptake was found more frequently with fat deposition (41/65 VQs, 63,1%) as compared to BME (35/65 VQs, 53,8%), while the combination of BME, fat deposition and increased signal in PET/MRI was found in 28/65 VQs (43,1%). A total of 20/65 VQs (29.4%) showed a combination of syndesmophytes on CT and increased signal in PET/MRI.
Conclusions The new technique of combining the PET and MRI provides largely similar imaging signals as conventional MRI. However, we did observe differences between the two techniques, especially in areas with less inflammatory activity but where bone metabolism might be more prominent, such as areas with (posti-inflammatory) fatty deposition or not yet fused syndesmophytes confirmed by CT. Especially the possibility of PET to detect osteoblastic activity in areas where no inflammatory signal is detected seems to be of interest for the understanding and prediction of structural disease progression in axial spondyloarthritis.
Disclosure of Interest None declared