Plain radiography in the evaluation of knee osteoarthritis.

The prospects for results of controlled studies of disease-modifying osteoarthritis drugs in humans have fostered interest in radiographic methods to detect early cartilage damage and to assess progressive cartilage changes in osteoarthritis of the knee. We summarize here the extent to which technical factors in plain knee radiography-ie, the degree of knee flexion, misalignment of the x-ray beam, magnification of the radiographic image of the joint-and mensural procedures can impact the reproducibility of quantitative measurements of tibiofemoral joint space width, the surrogate for thickness of articular cartilage. Failure to standardize crucial elements of conventional knee radiography and to computerize measurement of tibiofemoral joint space width will introduce significant (and probably insurmountable) error variation into radiographic outcome data. The level of precision achievable with fluoroscopically-assisted flexion of the knee and rotation of the foot with computerized, magnification-corrected measurement of joint space width greatly improves the feasibility of disease-modifying osteoarthritis drug trials as they relate to sample size and duration of treatment necessary to detect an effect from disease-modifying osteoarthritis drugs.