Aim: There is great variability in the structural damage presentations of knee osteoarthritis (OA), which may be one of the reasons for the limited success of interventions intended to modify the disease course so far. We aimed to identify distinct trajectories of change in structural pathology in knees at risk for OA up to four years before incident radiographic OA (ROA) and examine differences in OA risk factors.
Methods: We analyzed data from 336 knees from the Osteoarthritis Initiative that developed incident radiographic OA during follow-up. Knee magnetic resonance images (MRIs) up to four years before incident ROA were scored semi-quantitatively for effusion-synovitis, cartilage damage, meniscal morphology (including meniscal tear and maceration) and bone marrow lesions. Group-based multi-trajectory modeling was used to identify distinct patterns in the combined trajectories of these MRI features. Baseline characteristics including OA risk factors were compared between the trajectory-based subgroups.
Results: Four distinct subgroups were identified: 1) bone, cartilage and inflammation with low prevalence of meniscal lesions (31%); 2) prevalent meniscal lesions with early severe cartilage damage (29%); 3) early severe cartilage damage with subsequent bone changes (no meniscal lesions and limited inflammation) (14%); and 4) prevalent meniscal lesions and proportional involvement of other features (26%). Differences in risk factors were found, such as a greater prevalence of obese women in subgroup 1; a predominance of older males with a higher incidence of knee injury and family history of TKR in subgroup 2; and a greater prevalence of sedentary females and non-whites in subgroup 3 compared with all other groups.
Conclusions: These findings suggest that there are differences in the contribution of each articular tissue to OA initiation and that this may be associated with differences in OA risk factors. Strategies for prevention of OA development and progression may need to be tailored to each disease subgroup.