Summary

Blood flow signals (BFSs) through the bone cortex in Doppler mode on ultrasonography (US) significantly correlate with bone marrow lesions (BMLs) on MRI, suggesting that BFSs through the bone cortex on US can be used as a screening test for detecting BMLs in patients with early knee osteoarthritis.

Abstract

Introduction

Ultrasonography (US) has also recently been reported as a diagnostic tool for knee osteoarthritis (OA). Blood flow signals (BFSs) in Doppler mode on US can be used to evaluate fracture healing and assess local blood flow. This study focused on BFSs through the bone cortex in Doppler mode on US. There are few detailed pathological reports; however, BFSs have been observed in patients with knee OA and are thought to indicate the presence of bone lesions. Bone marrow lesions (BMLs) on magnetic resonance imaging (MRI) are characteristic imaging findings of knee OA. BMLs have been reported to reflect high metabolic activity of the bones and microfractures. BFSs through the bone cortex and BMLs can be used to assess bone lesions. However, no studies have reported their correlation. This study aimed to assess whether BFSs entering the bone on US can be used as a screening test to detect BMLs on MRI in patients with early knee OA.

Methods

This nonrandomized, prospective, multicenter clinical trial included patients with knee joint pain who were diagnosed with early knee OA based on the Luyten diagnostic criteria between January 2018 and April 2024. We enrolled 107 patients who underwent MRI and confirmed the presence or absence of BFSs in the bone cortex on US. An orthopedic surgeon at each center examined the patients in the supine position with slight knee flexion. Cases in which BFSs were detected along the long or short axis in power or color Doppler mode on US were defined as BFS-positive. Based on previous studies, the knee joint was then divided into 15 compartments based on the whole-organ MRI score, and the BML score was calculated for each area. Areas with a total score of ≥1 were defined as having BMLs. For MRI assessment, one orthopedic surgeon blinded to the US findings assessed the BMLs. The presence or absence of BFSs or BMLs was confirmed in each patient. The association between BFSs and BMLs was evaluated using the chi-square test. The sensitivity and specificity of BFS in detecting BMLs on MRI were then calculated. In addition, the characteristics of BFSs passing through the bone cortex on US were investigated. Statistical significance was set at p <0.05.

Results

There were 42 male and 65 female patients, with a mean age of 67.4 years. The chi-square test revealed that BFS and BMLs in the femur and tibia were significantly associated (femur: χ2 [1] = 73.9, p <0.001; tibia: χ2 [1] = 72.2, p <0.001). For the femur, the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of BFSs in detecting BMLs on MRI were 82.8%, 97.4%, 92.3%, and 93.8%, respectively. For the tibia, they were 87.5%, 94.7%, 87.5%, and 94.7%, respectively.

Conclusion

BFSs and BMLs exhibited a strong correlation with high sensitivity, specificity, PPV, and NPV. Therefore, BFSs through the bone cortex on US can be used as a screening test to detect BMLs on MRI in patients with early knee OA.