Background/Purpose: Radiographic joint space width (JSW) is frequently employed to indirectly assess structural progression in the medial femorotibial compartment, but it is unclear whether assessment of the lateral compartment provides accurate results. In contrast, magnetic resonance imaging (MRI) directly delineates articular cartilage and has been employed for quantitative analyses of cartilage thickness status and progression in both the medial and the lateral compartment. In the current study, we compared the sensitivity to change between recently proposed radiographic fixed-location measures in the lateral femorotibial compartment (obtained from fixed-flexion radiography [FFR]) – vs. MRI-based cartilage thickness measures, in knees with and without lateral JSN.

Methods:

One- and two-year change was assessed in one knee of 290 OAI participants with radiographic OA (210 female, age 61.8±9.1 y, BMI 29.5±4.9 kg/m²) based on FFR and DESS (n=161) or FLASH (n=129) MRI. Of the 290 knees, 147 had definite osteophytes only and 57/78/8 had grade 1/2/3 lateral JSN. Fixed-location JSW was measured semi-automatically at 9 locations in the lateral femorotibial compartment (LFTC) between 70% (JSW(X=0.7)) and 90% (JSW(X=0.9)) of the femoral width (medial to lateral). Cartilage thickness was measured by manual segmentation in the entire LFTC and in combined external (eLFTC), central (cLFTC), and internal (iLFTC) lateral subregions. The standardized response mean (SRM=mean change / standard deviation of change) was calculated as measure of the sensitivity to change. The most sensitive MRI and FFR measures were selected for comparing the sensitivity between MRI and FFR.

Potential sample sizes for a two-arm study with an estimated treatment effect of 50 % (i.e. reduction in structural progression) was calculated using G*Power based on sensitivity results of the current study. The calculation was based on the assumption of parametric t-tests, a false positive rate of 0.05, a false negative rate of 0.2, and 15% drop-out. 

Results:

The greatest overall sensitivity to change in both knees without JSN and knees with lateral JSN was observed in cLFTC for MRI and in JSW(X=0.8) for FFR. In knees without JSN, the one-year SRM was relatively low in both MRI (-0.16) and FFR (-0.11). This results in required sample sizes of n=6,165 (MRI) and 11,264 (FFR). Over two years, a greater SRM was observed for both MRI (-0.30) and FFR (-0.20), reducing the required sample size to 1,626 (MRI) and 3,502 (FFR).

In knees with lateral JSN, the one-year SRM observed for MRI (-0.73) was twice that for FFR (-0.37), suggesting sample sizes of 286 (MRI) and 1048 (FFR). Over two years, the SRM for MRI was -0.86 and that for FFR was -0.64, resulting in required sample sizes of 204 / 364 knees, respectively.

Conclusion: The study shows that fixed location measures of lateral JSW can be used to follow structural progression in the lateral femorotibial compartment in knees with lateral JSN. However, the sensitivity to change tended to be greater for MRI than for FFR, particularly over 1 year. Treatment studies with reasonable sample sizes appear feasible when following knees with lateral JSN using MRI over one, or using FFR over two years.

---

« Back to 2013 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/sensitivity-to-change-of-radiographic-fixed-location-and-magnetic-resonance-imaging-based-cartilage-thickness-measures-in-the-lateral-compartment-of-knees-with-radiographic-osteoarthritis-da/