Osteophytes and progression of knee osteoarthritis.

OBJECTIVES Osteophytes are thought to stabilize an osteoarthritic joint, thereby preventing structural progression. Meagre longitudinal data suggest, however, that they are associated with an increased risk of structural progression. Our objective was to evaluate the effect of osteophyte size on radiographic progression in osteoarthritis (OA). METHODS Using data from a natural history study of persons with symptomatic knee OA, we obtained fluoroscopically positioned postero-anterior (PA) radiographs at baseline, 15 and 30 months. Using an atlas, osteophyte size was scored on a scale of 0-3 at each of four sites on the PA film and, for each knee, both compartment-specific (i.e. medial; lateral) and overall osteophyte scores were computed. Progression was defined as an increase over follow-up in medial or lateral joint space narrowing, based on a semiquantitative grading. Mechanical alignment was assessed using long limb films at the 15 month examination. Logistic regression was used to evaluate the relation of osteophyte size with progression, adjusting for age, gender and body mass index, and with and without adjustment for alignment. RESULTS Of 270 subjects who had 470 eligible knees with follow-up, 104 (22%) knees showed progression. Overall, osteophyte score modestly increased the risk of progression [odds ratio (OR) per S.D. increase of osteophyte score=1.4 (95% CI 1.1, 1.8, P=0.02)], but this effect weakened and became non-significant after adjustment for limb alignment (OR=1.3). Compartment osteophyte score was strongly associated with malalignment to the side of the osteophyte (e.g. medial osteophyte and varus). Compartment-specific osteophyte score markedly increased the risk of ipsilateral progression (e.g. medial osteophytes --> medial progression) [OR per S.D.=1.9 (95% CI 1.5, 2.5, P<0.001)] and decreased the risk of contralateral progression [OR per S.D.= 0.6 (95% CI 0.5, 0.8, P=0.002)], but these associations diminished when we adjusted for limb alignment (OR=1.5 and 0.7 respectively). CONCLUSIONS Large osteophytes do not affect the risk of structural progression. They are strongly associated with malalignment to the side of the osteophyte, and any relation they have with progression is partly explained by the association of malalignment with progression.