Background: Excessive bone formation is an important hallmark of ankylosing spondylitis (AS). Recently, it has been demonstrated that axial bony lesions in AS patients can be visualized using [18F]fluoride PET-CT. This technique may also provide opportunities to monitor changes in bone formation in axial AS lesions during therapy. Objectives: To assess (1)changes in [18F]fluoride uptake in axial lesions of AS patients before and 12 weeks after treatment with TNF-inhibitors (TNFi) and (2)whether [18F]fluoride uptake on PET is related to focal bone formation in spine biopsies. Methods: Twelve TNFi naive AS patients (female 7/12; age 39+/-11years) with high disease activity (BASDAI 5.5+/-1.1) were included. [18F]fluoride PET-CT scans were performed before initiation of TNFi therapy. In 2 patients, biopsies were obtained from PET identified spine lesions for histologic analysis. Of the remaining 10 patients, a second [18F]fluoride PET-CT scan was performed after 12 weeks of TNFi treatment. PET scans were scored visually for positivity by two blinded expert readers. Subsequently, [18F]fluoride uptake was quantified in PET positive (PET+) lesions using the standardized uptake value corrected for individual integrated whole blood activity concentrations (SUVAUC). Clinical response to TNFi was defined according to ASAS20 at 24 weeks. Results: At baseline, in all patients at least one axial PET+ lesion was found. In spine, 6/10 patients showed 84 lesions (range 2-30;63% thoracic spine, primarily costovertebral joints) and in the SI joints in 9/10 patients were PET+ (Fig A). Histological analysis of PET+ lesions in the spine confirmed local osteoid formation, which was nearly absent in PET negative lesions. Quantitative PET analysis revealed significantly lower [18F]fluoride uptake in spine lesions at baseline in responders than in non-responders. This difference remained after 12 weeks of treatment (mean difference in SUVAUC: -0.5, 95% CI:[- 0.7,-0.2], p=0.001). After 12 weeks of TNFi treatment, [18F]fluoride uptake in clinical responders decreased significantly in the costovertebral (mean difference SUVAUC: -1.0, 95% CI: [-1.3,-0.7]) and SI joints (mean difference in SUVAUC: -1.2, 95% CI: [-2.3,-0.2]) (fig B) in contrast to non-responders (mean difference in SUVAUC: -0.4, 95% CI: [-2.3,1.6] and +0.4, 95% CI: [-0.6, 1.4], respectively). [18F]fluoride uptake in other spinal lesions such as bridging syndesmophytes showed heterogeneous response without a significant decrease in [18F]fluoride accumulation over time at a group level. Conclusions: [18F]fluoride PET-CT enables non-invasive visualization of (changes in) lesions with bone formation of the whole spine and SI joints in clinically active AS patients, which is confirmed by histological signs of osteoid formation. Part of these lesions, in particular costovertebral lesions in spine and SI joints, decreased in clinical responders to TNFi (and not in non-responders), whereas other spinal lesions remained unchanged at a group level. (Figure Presented).