Abstract
Background:
The etiology for recurrent patellar dislocations in children with patellofemoral instability (PFI) is not always apparent. Radiographic and other imaging measures help guide the clinician. However, the imaging parameters for pathologic pediatric PFI and the significance of these measures based on severity are not clearly defined. This study compares common imaging measures for PFI in a non-pathologic and known patellar dislocation pediatric population and determines the likelihood of symptoms based on the severity of each individual measure.
Methods:
An IRB-approved retrospective review of knee MRIs and radiographs at a single pediatric institution identified 108 limbs (102 patients) meeting inclusion criteria. Sixty-nine limbs had no known patellar pathology (control group) and 39 had surgery for recurrent patellar dislocations (surgical group). MRI measures included tibial tuberosity-trochlear groove distance (TT-TG), tibial tuberosity-posterior cruciate ligament distance (TT-PCL), and lateral trochlear inclination (LTI). Radiographic measures included the Caton-Deschamps Index (CDI) and Insall-Salvati Index (ISI). These measures are representative of PFI risk factors, specifically patellar tendon malalignment, trochlear dysplasia, and patella alta. Binomial logistic regression was used to relate the anatomical parameters to the probability of PFI.
Results:
The surgical and control groups were similar in terms of age, height and weight. The logistic regression analysis indicated that the TT-TG (p=0.001) and LTI (p<0.001) were correlated with PFI. The odds ratio (95% confidence interval) was 1.19 (1.07 -1.31) for TT-TG, giving the increased probability of a knee to experience PFI for each 1 mm increase. The odds ratio was 0.85 (0.77-0.93) for LTI, showing a decrease in the probability of instability with each 1° increase. Neither CDI nor ISI measures significantly correlated with probability of PFI (p=0.30 and p=0.18, respectively). The two parameters of TT-TG and LTI accounted for 42% (Nagelkerke R2) of the variance between the control and surgical groups.
Conclusion/Significance:
Comparing pediatric patients with recurrent patellar dislocations to controls, logistic regression analysis suggests the likelihood of PFI correlates with the severity of specific radiographic measures, specifically those representing the degree of patellar tendon alignment (TT-TG) and trochlear dysplasia (LTI). Knowing this correlation can help guide the treatment of PFI in pediatric patients by providing physicians and families with a continuum on which to base their decisions.