ASSESSING TEST-RETEST RELIABILITY OF JSW MEASUREMENTS FOR THE MOST4 KNEE RADIOGRAPH POSITIONING PROTOCOL

Abstract

INTRODUCTION

Despite challenges, joint space width (JSW) on weight-bearing radiographs has remained the most common structural outcome for Phase III trials approved by either the FDA or the EMA. The JSW is the distance between the projected femur and tibia margins on radiographic images. Superimposition of the posterior and anterior edges of the tibial plateau is required to accurately measure radiographic JSW. Knees must be positioned correctly at both baseline and follow-up to maximize reliability of measures of JSW and reliability metrics are useful to calculate sample sizes for studies that depend on JSW measurements.

OBJECTIVE

To establish the test-retest reliability of knee joint 2D imaging protocols used by the Multicenter Osteoarthritis Study (MOST4) at the 20-year visit.

METHODS

This cross-sectional, observational study compared joint space width (JSW) measurements between baseline and 2-week follow-up on PA radiographs acquired using an XFI scanner (Planmed Oy, Helsinki, Finland), using the same protocol used for the MOST4-V1 (20-year) visit. Participants were age ≥50 years, Kellgren-Lawrence grade 0-4, met MOST4 inclusion criteria, and could safely undergo WBCT (i.e. body mass< 230kg, able to stand for 5 minutes with support). Participants were positioned using a customized Perspex frame which could be tilted in the sagittal plane (5°, 10° or 15°) to make the tibial plateaus coplanar with the x-ray beam for a PA radiograph of both knees. Shrout-Fleiss intraclass correlation coefficients (ICC 2.1) and limits of agreement (mm) were calculated for joint space width (JSW) measurements on radiographs acquired at baseline and 2-week follow-up. For the medial compartment, JSW was measured at 50%, 60%, 70%, 80%, and 90% of the distance from the medial tibial spine (0%) to the outer margin of the medial tibial plateau (MTP; 100%). For the lateral compartment, JSW was measured at 50%, 60%, 70%, 80%, and 90% of the distance from the lateral tibial spine (0%) to the outer margin of the lateral tibial plateau (100%). JSW could not be measured in one knee due to poor alignment of the MTP and some knees could not be measured at JSW 90% due to the relative position of the femur on the tibia.

RESULTS

A total of 28 knees for 14 participants (demographics in Table 1) were imaged. ICC for JSW at each position in the medial and lateral compartments at JSW 50% to 90% are presented in Table 2 and limits of agreement are presented in Table 3. ICC for JSW measurements in the medial compartment were all ≥ 0.92 and ICC for JSW measurements in the lateral compartment were all ≥ 0.76.

CONCLUSION

These data support that the MOST4 knee radiograph acquisition protocol using the XFI has acceptable test-retest reliability for measurement of JSW at locations within the medial and lateral compartments.