Background Osteophytes are important signs in knee osteoarthritis as they are one of the diagnostic criteria and are important in assessing disease severity (Kellgren-Lawrence grading). Musculoskeletal ultrasound (MSUS) is being adopted by a growing number of rheumatologists because of its cost-effectiveness and the absence of radiation exposure.
Objectives The aim of the study was to evaluate the concordance of radiographs and MSUS in terms of detecting osteophytes in knee osteoarthritis.
Methods A cross-sectional study in outpatients presenting with knee osteoarthritis, according to the 1990 ACR criteria for knee osteoarthritis. Radiographs of the knees (anteroposterior incidence) and MSUS of the knee have been performed for each patient. Two rheumatologists trained in MSUS (S.S and A.H) have done MSUS examinations, and had no access to patients’ radiographs. Radiographs were sent and read by an independent rheumatologist (B.I) in another rheumatology center. Osteophytes were searched on 4 sites, chosen because of their easy access by both radiographs and MSUS: medial and lateral femoral condyles; medial and lateral superior tibial extremities. Osteophytes of the femoropatellar compartment have not been assessed because they require strict lateral radiographs and are not easily accessible on MSUS, thus a concordance assessment could not be reliable on this site. Concordance has been assessed using Kappa concordance coefficient.
Results We have examined 192 knees of 102 patients, 76.5% females, mean age 65.2 ± 9.7 years. Radiographic analysis allowed the detection of osteophytes in 35.9% cases, with a mean number of 1.6 osteophytes per knee when present, while MSUS detected osteophytes in 86.5% cases, with a mean number of 2.7 osteophytes per knee, when present. The Kappa coefficient between radiographs and MSUS was 0.183 (95% CI: 0,138-0,228). Most of the detected osteophytes on MSUS (60.5%) were located on the femoral condyles.
Conclusions Concordance between radiography and MSUS in the detection of osteophytes is poor, due to a much higher level of detection of MSUS (2.5 much more osteophytes detected with MSUS). This was made possible since new high resolutions machines were made available. The place of knee MSUS is still to be defined in the diagnostic procedure of knee pain, may be in case of normality of radiographs. To do this, studies assessing sensitivity/specificity of knee MSUS in detecting infra-radiographic knee osteoarthritis are required.
Disclosure of Interest None Declared