In patients with rheumatoid arthritis (RA), a novel visual scoring method to assess finger joint cortical interruptions using high-resolution peripheral quantitative computed tomography (HR-pQCT) detected small interruptions successfully and with decent reliability, despite suboptimal feasibility, according to findings published in PLoS One. Although interruptions were seen in patients with RA and healthy controls, they were more common and larger in patients with RA.
Bone microarchitecture can be readily visualized using HR-pQCT, making this technique ideal for identification and quantification of small, early bone changes. However, there had been no validated scoring methods proposed to evaluate cortical interruptions and vascular channels or erosions. Investigators sought to develop a scoring system that would be useful in distinguishing physiologic from pathologic morphology.
There were 10 healthy female control patients (mean age, 46.1 years) and 20 female patients with RA (mean age, 57.4 years; mean disease duration, 117.1 months) selected from the MOSA-Hand study for this analysis. In both groups, cortical interruptions and adjacent trabecular disruptions in the second and third metacarpophalangeal (MCP) and proximal interphalangeal (PIP) joints were assessed by 2 independent readers blinded to clinical and demographic information. Intra- and inter-reader reliabilities were calculated using intraclass correlation coefficients (ICC), along with feasibility. Expert discussions, consensus meetings, reading experiments, and a literature review contributed to the development of the scoring method.
In a total of 98 joints, researchers detected 252 cortical interruptions (214 in patients with RA and 38 in healthy control patients), of which 21% had associated trabecular distortion (all in patients with RA). The majority were seen on at least 2 consecutive slices. Individuals with RA demonstrated significantly wider mean interruption diameters vs controls (0.88 vs 0.47 mm; P =.03). Patients with RA had more interruptions in the radial and ulnar quadrants compared with controls (P <.05). The overall mean reading time per joint was 9.2 minutes.
Combined mean intra-reader reliability resulted in an ICC of 0.88 (95% CI, 0.83-0.92) per joint and an ICC of 0.69 (95% CI, 0.65-0.73) per quadrant. Inter-reader reliability showed an ICC of 0.48 (95% CI, 0.20-0.67) per joint and 0.56 (95% CI, 0.49-0.62) per quadrant, respectively. There was moderate intra-reader reliability (ICC, 0.52-0.67) for interruption presence, but excellent intra-reader reliability (ICC, 0.69-0.88) for cortical interruption number. Inter-reader reliability for both presence and number was fair to moderate (ICC, 0.37-0.56).
As a limitation of the study, the researchers noted that adjacent trabecular distortions underestimate damage extent as a consequence of the exclusion of completely destroyed bone ends.
“Our visual scoring method incorporates all cortical interruptions in MCP and PIP joints and hereby allows to make a distinction between small cortical interruptions and larger cortical interruptions which are considered more specific for RA,” observed the authors, noting that while extended scoring times per joint may limit feasibility in clinical practice, the method would still be useful for future longitudinal clinical studies.
Disclosures: This study was supported by the Weijerhorst foundation and Pfizer. Bert van Rietbergen is a consultant for Scanco Medical AG. The other authors declare no conflicts of interest.
Scharmga A, Peters M, Bergh JPVD, et al. Development of a scoring method to visually score cortical interruptions on high-resolution peripheral quantitative computed tomography in rheumatoid arthritis and healthy controls. PloS One. 2018;13(7):1-13.