Patellar Tilt and Patellar Tendon–Trochlear Groove Angle Present the Optimum Magnetic Resonance Imaging Diagnostic Reliability for Patients With Patellar Instability

Abstract

To describe, in controls and in a population with patellar instability, magnetic resonance imaging values of measurements representing major associated factors for patellar instability (patellar height, trochlear dysplasia, and extensor mechanism alignment), as well as their cutoff values. In total, 323 knee magnetic resonance imaging scans, 142 with patellar instability and 181 controls without patellofemoral complaints (anterior cruciate, medial collateral ligament, meniscus ruptures or normal) were evaluated. Means, normality values in the control population, ideal cutoff values through receiver operating characteristic curves analysis, and interobserver reliability (intraclass correlation coefficient) were described for a series of measurements. All measurements were statistically different in control and instability patients, except for the patellotrochlear index and tibial tuberosity to posterior cruciate ligament distance. The interobserver intraclass correlation coefficient was good or excellent (above 0.75) only for the patellotrochlear index, patellar tendon-trochlear groove (PTTG) angle, and patellar tilt. The optimal cutoff value for each measurement was: PTTG angle ≥25.3o with sensitivity (S) of 70% and specificity (E) of 89%, patellar tilt ≥16o (S: 69% and E: 84%), trochlear sulcus angle ≥153o (S: 75% and E: 76%), Carrillon ≤12.8o (S: 62% and E: 87%), PTTG distance ≥11mm (S: 71% and E:78%), Caton-Deschamps index ≥1.23 (S: 72% and E: 76%) and trochlear bump ≥3.95 mm (S: 76% and E: 65%). Caton-Deschamps index (≥1.23), trochlear sulcus angle (≥153o), ventral prominence of the trochlea (≥3.95 mm), PTTG distance (≥11 mm), PTTG angle (≥25.3o), Carrillon angle (≤12.8o), and patellar tilt (≥16o) presented better diagnostic performance for patellar instability. Patellotrochlear index and tibial tuberosity to posterior cruciate ligament distance were not related to patellar instability. The interobserver reliability of the factors related to patellar instability was excellent only for the PTTG angle and lateral patellar tilt. III, retrospective case-control study.