Development of CT-based methods for longitudinal analyses of paranasal sinus osteitis in granulomatosis with polyangiitis

Background Even though progressive rhinosinusitis with osteitis is a major clinical problem in granulomatosis with polyangiitis (GPA), there are no studies on how GPA-related osteitis develops over time, and no quantitative methods for longitudinal assessment. Here, we aimed to identify simple and robust CT-based methods for capture and quantification of time-dependent changes in GPA-related paranasal sinus osteitis and compare performance of the methods under study in a largely unselected GPA cohort. Methods GPA patients (n = 121) with ≥3 paranasal CT scans obtained ≥12 months apart and control patients not having GPA or rhinosinusitis (n = 15) were analysed by: (i) Global osteitis scoring scale (GOSS), originally developed for chronic rhinosinusitis; (ii) Paranasal sinus volume by manual segmentation; (iii) Mean maxillary and sphenoid diameter normalised to landmark distances (i.e. diameter ratio measurement, DRM). Results Time-dependent changes in GPA-related osteitis were equally well measured by the simple DRM and the labour-intensive volume method while GOSS missed ongoing changes in cases with extensive osteitis. GOSS at last CT combined with DRM identified three distinct patient groups: (i) The no osteitis group, who had no osteitis and no change in DRM from baseline CT to last CT (45/121 GPA patients and 15/15 disease controls); (ii) Stable osteitis group, with presence of osteitis, but no change in DRM across time (31 GPA); (iii) Progressive osteitis, defined by declining DRM (45 GPA). Conclusions We suggest DRM and GOSS as complementary methods for capturing, classifying and quantifying time-dependent changes in GPA-related osteitis.