CALCIUM CRYSTAL DEPOSITION AND KNEE OSTEOARTHRITIS, ASSESSMENT OF JOINT INFLAMMATION BY DCE-MRI: A CROSS-SECTIONAL STUDY

Abstract

INTRODUCTION

Calcium crystal (CaC) depositions in hyaline cartilage, meniscus and the joint capsule are seen in some patients with knee OA. Whether or not these crystals exacerbate the symptoms and progression of OA is not well understood. Composed primarily of calcium pyrophosphate and basic calcium phosphate crystals, CaC has been shown to activate pro-inflammatory pathways in vitro. The pro-inflammatory effect of these crystals in vivo is more uncertain.

OBJECTIVE

In this exploratory cross-sectional analysis we aimed to investigate if overweight individuals with knee OA and CaC deposits experience more knee joint inflammation compared with matched individuals without CaC deposits.

METHODS

We used pre-randomization imagining data from an RCT, the LOSEIT trial. Participants were included if they were between 18 and 75 years old; had clinical knee OA, according to the ACR criteria; showed KLG 1-3 on weight-bearing x-ray; and had a BMI ≥ 27 kg/m². Participants had CT (Somatom Definition Edge®, Siemens, Germany) and 3T MRI (Verio®, Siemens, Germany) of the index knee. Intraarticular CaCs were assessed on CT (in-plane resolution: 0.6 × 0.6mm, slice thickness: 1mm, tube voltage: 140 kV) using a modified version of the Boston University Calcium Knee Score (BUCKS), classifying participants as OA with CaC if they had a BUCKS ≥ 1 in any sub-region. To estimate joint inflammation, we used both static and dynamic contrast-enhanced (DCE) MRI. The following static MRI variables were analyzed: MRI in OA Knee Score (MOAKS) with Hoffa-synovitis and effusion-synovitis scores summed to one MOAKS-synovitis score (0–6). The Boston-Leeds Osteoarthritis Knee Score (BLOKS) effusion sub-score (0–3) and the 11-point whole-knee synovitis score (CE-synovitis) as proposed by Guermazi et al. (0–22). Heuristic DCE-MRI analysis was carried out using the software Dynamika® v. 5.2.2 (Image Analysis Group). We included five DCE-MRI variables; Initial Rate of Enhancement (IRE), Maximum Enhancement (ME), Most Perfused Voxels (Nvoxel) and the two composite scores; IRE x Nvoxel and ME x Nvoxel. We only included participants with complete CT and MRI data, i.e., no imputation for missing data. To test if there was a difference in the MRI variables between participants with and without CaC deposits, we used an Analysis of Covariance (ANCOVA) model adjusted for age and KLG. We did not adjust for multiple testing, acknowledging the exploratory nature of this study and interpreting the results accordingly.

RESULTS

Of the 168 participants included in the LOSEIT trial 115 had MRI available; 13 (11.3 %) had CaC deposits, 8 in the cartilage, 5 in the meniscus and 2 in the joint capsule. Mean (SD) static and DCE-MRI variables are presented in Table 1 along with the results from the ANCOVA analyses. None of the MRI variables were associated with the presence of CaC deposits (Figure 1). The between-group differences were small for all MRI variables and the standardized mean differences ranged from small to medium (0.31-0.56) (Table 1).

CONCLUSION

In individuals with knee OA, we did not find an association between intraarticular CaC deposits and an increase in knee joint inflammation assessed by static and DCE-MRI.