Profiling of knee osteoarthritis patients and progression to total knee arthroplasty based on symptom trajectories: Evidence on more than 6000 knees from the Osteoarthritis Initiative.

Introduction This study aimed at investigating the symptomatic evolution of knee osteoarthritis (OA) and identifying patient clusters based on symptom trajectories up to the need for total knee arthroplasty (TKA). Methods Patients were selected from the Osteoarthritis Initiative database if presenting complete data for baseline age, sex, BMI, Kellgren-Lawrence (KL) grade, knee effusion, knee extensor mechanism tendinopathy, and WOMAC scores at baseline and each annual follow-up for 8 years. Patients were clustered based on their longitudinal WOMAC trajectories. The resulting groups were characterised according to the multivariate analysis results on baseline factors. A WOMAC literature cutoff of 28.8 points was used as a pragmatic reference for TKA indication. Results 6,484 knees were included. Multivariate analysis showed that the main baseline factors influencing WOMAC trajectories were KL grade, female sex, knee effusion, and extensor mechanism tendinopathy (all p<0.05). Patients clustering identified seven patient groups according to the level and shape of their longitudinal WOMAC trajectories. Based on the TKA literature cutoff of 28.8 points, three macro-groups were identified: patients not requiring, at risk of, or likely requiring TKA in the following 8 years. Conclusions The clinical evolution of knee OA up to TKA is influenced by several patient characteristics, including KL grade, female sex, knee effusion, and extensor mechanism tendinopathy. Seven patient groups were identified by clustering longitudinal WOMAC trajectories, and three macro-groups were defined according to the progression to TKA indication. These results provide a pragmatic framework to support risk stratification and more tailored care for knee OA patients in clinical practice.