Patients with knee osteoarthritis can be divided into subgroups based on tibiofemoral joint kinematics of gait - an exploratory and dynamic radiostereometric study

Abstract

Objective: Patients with advanced knee osteoarthritis (KOA) frequently alter their gait patterns in an attempt to alleviate symptoms. Understanding the underlying pathomechanics and identifying KOA phenotypes are essential to improve treatments. We investigated kinematics in patients with KOA to identify subgroups of homogeneous knee joint kinematics.

Method: A total of 66 patients with symptomatic KOA scheduled for total knee arthroplasty and 15 age-matched healthy volunteers with asymptomatic, non-arthritic knees were included. We used k-means clustering to divide patients into subgroups based on dynamic radiostereometry-assessed tibiofemoral joint kinematics. Clinical characteristics such as knee ligament lesions and KOA scores were graded by magnetic resonance imaging and radiographs, respectively.

Results: We identified four clusters that were supported by clinical characteristics. The flexion group (n = 20) consisted primarily of patients with medial KOA. The abduction group (n = 17) consisted primarily of patients with lateral KOA. The anterior draw group (n = 10) was composed of patients with medial KOA, some degree of anterior cruciate ligament lesion and the highest KOA score. The external rotation group (n = 19) primarily included patients with medial collateral and posterior cruciate ligament lesions.

Conclusion: Based on tibiofemoral gait patterns, patients with advanced KOA can be divided into four subgroups with specific clinical characteristics and different KOA-affected compartments. The findings add to our understanding of how knee kinematics may affect the patient's development of different types of KOA. This may inspire improved and more patient-specific treatment strategies in the future.

Keywords: Clustering; Gait analysis; Kinematics; Knee osteoarthritis; Radiostereometry; Statistical parametric mapping.