Comparison of Marker-Based RSA and CT-RSA for Analyzing Micromotions After Distal Radius Osteotomy: A 1-Year Retrospective Study of 24 Patients

Abstract

Radiostereometric Analysis (RSA) is the most accurate method for determining early micromotions of orthopedic implants. Computed Tomography Radiostereometric Analysis (CT-RSA) is a method that can be used to determine implant and bone micromovements using low-dose CT scans. This study aimed to evaluate the reliability of the CT-RSA method in measuring the interfragmental mobility in patients who have undergone a correction osteotomy due to a malunited distal radius fracture. Twenty-four patients were included and operated with a radiolucent volar plate. Markers were embedded in the plate and bone. RSA and CT examinations were obtained postoperatively up to 1-year postoperative. Micromovements of the distal radius segment relative to the proximal were compared between the methods with paired analysis and Bland-Altman plots. The limits of clinical significance were: dorsal/volar tilt < 10°, radial shortening < 5 mm, radial inclination ≥ 15°, and radial shift < 5 mm. For the dorsal/volar tilt, the paired analysis between the two methods, showed a mean difference (95% CI) of -0.06° (-0.67 to 0.55), for radial compression-0.04 mm (-0.09 to 0.01), for radial inclination 0.21° (-0.06 to 0.48), and for radial shift -0.07 mm (-0.21 to 0.07). The paired analysis for micromotions showed that the thresholds of clinical significance are excluded from the difference's 95% CI. The Bland-Altman plots showed comparable results up to 1 year, considering clinically relevant thresholds. In conclusion, the CT-RSA method is comparable to that of marker-based RSA in measuring micromotions after wrist osteotomy, as the differences between the methods are not clinically significant.

Keywords: CT‐based; computed tomography; distal radius; micromotions; radiostereometric analysis.

Figure 1

Axes of interest in distal radius displacement. The dorsal/volar tilt corresponds to rotations around the x‐axis, radial inclination to rotations around the z‐axis, axial compression to translations along the y‐axis, and radial shift to translations along the x‐axis.

Figure 2

Flowchart showing inclusion of patients in the current study.

Figure 3

Registration of the tantalum markers of the proximal radius bone segment as reference object on two different CT scans (3.1A and 3.1B) and verification of the overlap process with the help of chromatic overlay (C). Registration of the tantalum markers of the distal radius bone segment as moving body on two different CT scans (2A and 2B) and verification of the overlap process with the help of chromatic overlay (2C). Definitions of the measurement point (3). Definition and adjustment of the coordinate system (4).

Figure 4

(A–D) Bland–Altman plot for translations on the three orthogonal axes as well as total translation. Limits of Agreement are shown as dotted, red lines with 95% confidence intervals in light blue. Bias is shown as a solid, blue line with the 95% confidence interval in green. The line of equality is shown as a dotted, orange line. (E–H) Bland–Altman plot for rotations around the three orthogonal axes as well as total rotation. Limits of Agreement are shown as dotted, red lines with 95% confidence intervals in light blue. Bias is shown as a solid, blue line with the 95% confidence interval in green. The line of equality is shown as a dotted, orange line.

Figure 5

(A–F) Bland–Altman plots of the measurements performed with the CT‐RSA method with coordinate system adjusted to the anatomy definition RSA strived for, versus the actual placement of each patient's arm. Limits of Agreement are shown as dotted, red lines with 95% confidence intervals in light blue. Bias is shown as a solid, blue line with the 95% confidence interval in green. The line of equality is shown as a dotted, orange line. The two outliers on all of the plots represent the same two patients.