Impact of contralateral radiographic status on longitudinal change of cartilage transverse relaxation time (T2) in radiographically normal knees, a model of early OA? – data from the osteoarthritis initiative

Abstract

Purpose: Evaluating treatment for modifying structural progression at an early stage of OA requires models that display subtle synovial tissue alterations that occur early in the disease process, but still have a reasonable likelihood of progression. We have previously shown that radiographically normal (KLG0) knees with definite contralateral (CL) radiographic OA had a greater likelihood of incident ROA, and that KLG0 knees with CL radiographic joint space narrowing (JSN) displayed greater cartilage loss than those with radiographically normal CL knees. The purpose of this study was to use this model of “early” OA of a KLG0 case knee in a person with CL (non-posttraumatic) ROA, to determine whether baseline values and longitudinal change in cartilage composition differ compared with those with radiographically normal CL knees (controls). Methods: The data were obtained from the Osteoarthritis Initiative (OAI). 150 KLG0 case knees displayed CL ROA (KLG2-4; central readings vs.0.7), without reporting previous injury to that knee. Of those, 63 (50.8% female, age: 64.2±8.8 y, BMI: 27.6±4.4 kg/m2) were right knees with multi-echo spin-echo (MESE) MRIs (Fig.1) to support analysis of the cartilage transverse relaxation time (T2), a method previously shown to be related to cartilage composition and mechanical properties. 23 were CL KLG2 without JSN, 22 KLG2 with JSN grade1, 13 KLG3, and 5 KLG 4. The 63 case knees were matched 1:1 with 63 KLG0 controls of persons who were bilaterally KLG0 and had the same sex, similar age (64.3±8.7 y), and BMI (27.3±4.0 kg/m2). The 3-year change in cartilage T2 (between year 1 and 4 [Y1/4] follow-up) was determined from sagittal MESE MRIs (Fig.), analyzing the superficial and deep 50% of the cartilage in each of 16 femorotibial joint (FTJ) subregions. The mean T2-values across the superficial and deep cartilage layers of the entire FTJ were the primary analytic focus. Exploratory measures included T2 of the medial and lateral femorotibial compartment (MFTC/LFTC) and location-independent T2 shortening and lengthening scores (i.e. the sum of all negative / positive femorotibial subregion T2 changes). Results: The baseline T2 (Y1) was significantly greater in the case than in control KLG0 knees, both in the superficial and deep FTJ cartilage layers (paired t-test: p=0.01; Cohen’s D=0.40; and p=0.04; Cohen’s D=0.34, respectively). In the superficial layer, the effect tended to be stronger in the medial than in the lateral, and in the deep layer stronger in the lateral than in the medial compartment (Table 1).Table 1Cartilage T2 at the beginning of the observation intervalCase KLG0 kneesControl KLG0 kneesDifferenceSuperficial LayerMeanSD95% CIMeanSD95% CIPCohen DFTJ48.73.447.849.547.42.846.748.10.010.40MFTC49.24.248.150.347.73.746.848.60.020.37LFTC48.13.747.249.147.13.046.347.90.050.31Deep LayerFTJ37.72.337.238.337.02.036.537.50.040.34MFTC38.22.637.538.937.82.537.238.40.320.16LFTC37.32.636.637.936.22.235.736.80.010.44 Open table in a new tab When restricting analysis to those with CL JSN, the effect size (Cohen’s D) was 0.49 in the superficial and to 0.36 in the deep layer (data not shown). The longitudinal increase in medial and lateral T2 from Y1 to Y4 of the case knees was significantly different from zero; it also tended to be greater in case than in control KLG0 knees, but the case-control differences did not reach statistical significance (data not shown). Only when restricting the analysis to those with CL JSN, we observed a significantly (p<0.05) greater increase in deep layer T2 in the case than in the control knees in the FTJ, and also in the lateral compartment and location-independent lengthening score (Cohen’s D 0.45-0.46). No significant differences were observed in the longitudinal changes of the superficial layer (Table 2).Table 2Longitudinal change in T2 over 3 years of observationKLG0 with CL JSNKLG0 with CL KLG0DifferenceSuperficial LayerMeanSD95% CIMeanSD95% CIPCohen DFTJ1.01.90.41.60.91.50.41.40.850.04MFTC1.42.50.62.21.42.20.72.10.950.01LFTC0.62.2-0.21.30.41.9-0.21.00.820.06Decrease-15.316.1-20.4-10.1-15.110.2-18.4-11.80.96-0.01Increase31.717.726.037.429.717.824.035.40.610.11Deep LayerFTJ1.12.50.31.90.11.7-0.40.70.040.46MFTC1.33.30.22.40.32.4-0.41.10.140.34LFTC0.92.50.11.7-0.11.8-0.60.50.040.45Decrease-13.413.5-17.7-9.0-18.217.0-23.7-12.80.150.32Increase32.329.522.941.721.812.717.725.80.050.46 Open table in a new tab The increase in deep layer T2 was significantly correlated with location-independent measures of cartilage thickness loss in the FT joint (r=-0.27, p<0.01). Conclusions: This is the first study to explore whether cartilage T2 is sensitive to identifying compositional changes in cartilage in a model of “early” knee OA, i.e. a KLG0 knee with a CL non-posttraumatic ROA knee. Significant differences in superficial and deep layer cartilage T2 were observed at the beginning of the longitudinal observation interval. Similar to previous observations for the change in location-independent measures of cartilage thickness, significant longitudinal differences in T2 lengthening between case and control knees were observed in knees with CL JSN. These differences were limited to the deep layer, and were most prominent in the lateral femorotibial compartment. KLG0 knees with CL JSN hence may represent a model of early OA in which structure modification can be studied at an early disease stage.