Quantitative Variable Assessment of Patellar Instability: An MRI-Based Study.

ABSTRACT

OBJECTIVE. The purpose of this study was to investigate the reproducibility of three quantitative MRI parameters associated with patellar instability and to determine whether they measure anatomic predisposition to patellar instability individually or in combination with the other parameters. MATERIALS AND METHODS. In this retrospective study, 100 patients diagnosed with a patellar dislocation injury and 100 age- and sex-matched control patients were examined using MRI. The distance between the tibial tubercle and posterior cruciate ligament (TT-PCL), distance between the tibial tubercle and trochlear groove (TT-TG), and TG depth (trochlear dysplasia) were measured independently by three fellowship-trained musculoskeletal radiologists. Intraclass correlation coefficient (ICC) was used to assess intraobserver and interobserver reliability. The parameters in both groups were tested for interdependence on each other and were compared for prevalence and association with patellar instability. RESULTS. All three parameters showed almost perfect intraobserver (TT-PCL ICC, ≥ 0.88; TT-TG ICC, 0.96; trochlear dysplasia ICC, ≥ 0.92) and interobserver (TT-PCL ICC, 0.82; TT-TG ICC, 0.94; trochlear dysplasia ICC, 0.91) reliability and were significantly more common in the patellar instability group. Trochlear dysplasia had the highest association with patellar instability, both as a unique parameter and in pairwise combination with an abnormal TT-TG. Optimal cutoff thresholds for normal TT-TG and TT-PCL were 15.00 mm or less and 21.30 mm or less, respectively. The optimal normal cutoff threshold for evaluating trochlear dysplasia via trochlear depth was 4.95 mm or more. CONCLUSION. Patellar instability is multifactorial. Highly reproducible parameters derived from MRI reveal both unique and overlapping anatomic predispositions, and considering all parameters together may help individualize patient management when selecting orthopedic procedures.