Objectives. Ultrasound (US) can detect cortical bone lesions in RA. However, not all cortical bone lesions are erosions.Herein, we aimed to define whether US can differentiate between physiological bone channels and pathological erosions in RA and to provide topographic description of their differential localization.
Methods. RA patients and healthy controls (HC) received US examination of the metacarpophalangeal (MCPJ)
and proximal inter-phalangeal (PIPJ) joints adjudicating cortical bone lesions as physiological bone channels or
pathological erosions. In a subset of RA patients and HC, high-resolution peripheral quantitative computed tomography (HR-pQCT) of the hand was performed to validate the classification of lesions.
Results. A total of 40 RA patients and 43 HC were enrolled and totally 771 MCPJ and 638 PIPJ were examined
by US, and 94 and 51, respectively, by HR-pQCT. US-defined cortical bone lesions clustered in the lateral part of the MCP (50%) and the dorsal part of the PIPJ (66.7%) in RA. US-defined physiological bone channels clustered
in the palmar parts of the MCPJ and PIPJ in both RA (78.8% and 100%, respectively) and HC (51.8% and 80%,
respectively). HR-pQCT data confirmed US data with respect to adjudication of physiological bone channels and
pathological erosions. Erosions were significantly (all P <0.000001) larger than physiological channels and preferentially localized at radial and ulnar sites, while physiological channels were clustered at palmar sites. Specificity of US was excellent for erosions in RA and for physiological bone channels in HC and RA.
Conclusion. US allows differentiation between physiological channels and bone erosions in RA.

  Rheumatology 2020;59:3784–3792