DEFINING TROCHLEAR DYSPLASIA VIA THE LATERAL TROCHLEAR INCLINATION ANGLE

BACKGROUND: Trochlear dysplasia (TD) is the principle anatomic risk factor for patellar instability (PI). The lateral trochlear inclination (LTI) is a quantitative measure of trochlear morphology on axial magnetic resonance imaging (MRI). Threshold values for patients at risk for PI via LTI measurement have ranged from 11(o) to 17(o). Our group recently described a new technique for characterizing LTI by referencing the fully-formed posterior femoral condyles. This novel measurement technique revealed that the historically-used method was comparable in reliability but significantly underestimated TD. PURPOSE: The purpose of this study was to define, with high specificity, a LTI value cutoff value consistent with patellar instability using the new LTI measurement technique. METHODS: MRI scans of 95 patients aged 9 to 18 years treated for PI at our tertiary referral center were compared with a control cohort of 98 age- and gender- matched patients with knee imaging but no clinical evidence of PI. LTI was measured as the angle formed between a line subtended from the cartilaginous surface of the lateral trochlea at its most proximal extent and a line parallel to the fully-formed posterior condyles. Receiver operator characteristic (ROC) curve analysis was conducted to establish a cutoff value with optimal specificity and sensitivity. RESULTS: Average LTI was significantly lower in the study group (5.9±10.4(o)) than control (18.9±5.8(o)) (p<0.001). Area under the ROC curve was 0.86. A proposed 8.9(o) LTI threshold angle achieves good discrimination between the two groups, with specificity of 0.904 and sensitivity of 0.684. CONCLUSION: Re-examination of previously described threshold values using ROC curve analysis of LTI measured in reference to the posterior condyles found 8.9(o) as the optimal threshold value to achieve good discrimination between patients with and without PI. The novel technique has previously been established to have excellent intra- and inter-observer reliability and produced significantly lower LTI values than the historically-used technique. Taken in concert, this new cutoff is valuable to improve future clinical decision making in regards to risk stratification, treatment algorithms and research purposes.