Predicting Risk of Recurrent Patellofemoral Instability With Measurements of Extensor Mechanism Containment

Background: Identifying risk factors for recurrent patellar dislocation after a primary dislocation may help guide initial treatment. Magnetic resonance imaging (MRI) measurements relating the alignment of the extensor mechanism to trochlear morphology have been shown to distinguish patients with dislocations from controls, but their usefulness in predicting the risk of a second dislocation is not known. Purpose: To identify the association of novel MRI measures of patellar containment with recurrent instability in pediatric patients presenting with a first-time patellar dislocation. Study Design: Cohort study (Prognosis); Level of evidence, 3. Methods: The study was conducted at a tertiary care children's hospital (2005-2014) on patients (age, 8-19 years) with a first-time patellar dislocation. MRI measurements were made by 2 independent raters. Interobserver reliability was assessed for all measurements via an intraclass correlation coefficient (ICC). Only measurements with an ICC >0.8 were included. Univariable and multivariable logistic regression analyses were used to evaluate variables associated with recurrence. Results: A total of 165 patients with a median age of 14 years and a slight (57.6%) female predominance was identified. The median follow-up length of the whole cohort was 12.2 months (interquartile range, 1.6-37.1 months). Subsequent instability was documented in 98 patients (59.4%). MRI measurements with excellent correlation (ICC > 0.8) were the tibial tubercle to trochlear groove distance (TT-TG), the tangential axial width of the patella, the tangential axial trochlear width, the axial width of the patellar tendon beyond the lateral trochlear ridge (LTR), and the tibial tubercle to LTR distance. In univariate analysis, all mentioned MRI measurements had significant association with recurrent instability. However, after both backward and forward stepwise regression analyses, the tibial tubercle to LTR distance was the only independent predictor of recurrent instability (P = .003 in both). Patients with a tibial tubercle to LTR distance value greater than -1 mm had a significantly higher rate of recurrent patellar dislocation (72%). Conclusion: Of numerous axial view MRI parameters, only the tibial tubercle to LTR distance demonstrated a statistically significant association with recurrent patellar instability upon multivariable logistic regression analysis during short-term follow-up of a pediatric population presenting with initial lateral patellar dislocation. Interobserver correlation of the tibial tubercle to LTR distance was good (ICC > 0.8) and similar to that of TT-TG.