Background High-resolution peripheral quantitative computed tomography (HR-pQCT) is a novel imaging instrument for bony damage in rheumatoid arthritis (RA).
Objectives Agreement on a case definition for pathologic erosions is required given the sensitivity of HR-pQCT for detecting cortical bone disruptions. The reliability of erosion detection and measurement between readers is crucial to evaluate with this new technology, as HR-pQCT is undergoing validation as an outcome measurement tool.
Methods HR-pQCT images of the 2nd and 3rd metacarpophalangeal joints of RA patients and control subjects were used in an iterative process to achieve consensus on a case definition for erosions. This case definition was applied by 11 independent readers to score 58 joints. Each surface (radial, ulnar, palmar, and dorsal) of the proximal phalanx and metacarpal head were characterized for image quality, presence of cortical break as well as appearance of the respective cortical break (physiological or pathological). Moreover, the total amount of pathologic erosions per quadrant as well as their width and depth in two perpendicular muliplanar reformations were recorded.
Results The case definition requires the presence of a cortical break extending over at least two slices and seen in 2 orthogonal views. The break must be non-linear in shape to differentiate from vascular channels penetrating the cortical surface, and have underlying trabecular bone loss. Inter-reader agreement for erosion detection was at 90.2%. Discrepant scoring was attributable to lower quality images with multiple bony pathologies overlapping each other. Erosion size ranged from 0.016 to 0.890 cm in maximal width (mean 0.184 cm) and 0.030 to 0.800 cm in maximal depth (mean 0.186 cm), with between 12.3 and 24.0% variability in measurement between readers.
Conclusions We have devised a new case definition for erosions visualized with HR-pQCT imaging. Inter-reader reliability for erosion detection is good, yielding promise to use HR-pQCT as an outcome measurement tool for bony damage. Further refinement of selection of landmarks for erosion size measurement, or automated volumetric methods, will be pursued.
Disclosure of Interest None declared