The crossing-sign angle and trochlear dysplasia length: new radiographic parameters to measure the dysplastic trochlea.

Abstract

Trochlear dysplasia is one of the most relevant anatomic factors predisposing patients to patellofemoral instability and its recurrence. This study introduces two novel radiographic measurements: the crossing-sign angle and the trochlear dysplasia length. A concordance study was conducted in patients aged 10-40 years with at least one episode of patellofemoral dislocation and radiographic evidence of trochlear dysplasia on a true lateral knee radiograph. Three observers independently measured the crossing-sign angle, the trochlear dysplasia length, and bump height, as well as classified the degree of trochlear dysplasia according to a modified Dejour classification. Inter- and intra-observer reliability for continuous variables were assessed using the intra-class correlation coefficient (ICC), while agreement for the modified Dejour classification was evaluated using Krippendorff's alpha (α) and weighted kappa (κ). Pearson correlation analysis was performed between the radiographic measurements, and analysis of variance (ANOVA) test between these and Dejour grades. Crossing sign angle demonstrated moderate inter-observer reliability (ICC = 0.71) and excellent intra-observer reliability (ICC = 0.96). Trochlear dysplasia length showed moderate inter-observer reliability (ICC = 0.73) and excellent intra-observer reliability (ICC = 0.95). Bump height exhibited good inter-observer reliability (ICC = 0.87) and excellent intra-observer reliability (ICC = 0.97). The modified Dejour classification demonstrated substantial inter-observer agreement (α = 0.62) and perfect intra-observer agreement (κ=1.00). Strong positive correlation was found between trochlear dysplasia length and bump height (r = 0.73; 95% CI 0.52 to 0.85; p < 0.001). Bump height and trochlear dysplasia length increased progressively from Dejour type A to higher grades (B and D, ANOVA p <0.001). Crossing sign angle and trochlear dysplasia length provide a novel quantification of trochlear dysplasia and demonstrate moderate inter- and excellent intra-observer reliability. Bump height and the modified Dejour classification, when using a standardized protocol, also showed high inter- and intra-observer reliability. Bump height and trochlear dysplasia length have strong positive correlation. The clinical and surgical implications of these new parameters need further investigation in future studies. II.