Abstract
Background:
The etiology of patellofemoral osteochondritis dissecans (OCD) is unclear. Biomechanical factors, such as abnormal patellofemoral morphology seen in patients with patellofemoral instability (PFI), may contribute to the development of patellofemoral OCD. The purpose of this study was to determine whether anatomic risk factors associated with PFI are also present in patients with patellofemoral OCD.
Hypothesis:
Patients with patellofemoral OCD will have knee morphology more similar to patients with PFI than healthy controls.
Methods:
This was a retrospective, case-control study comparing three groups of pediatric (≤19 years of age) patients: those with OCD of the patella or trochlea and no history of PFI; those with recurrent (≥2 dislocations) PFI with no evidence of OCD; and those with normal knee MRIs (controls). One knee per patient was included. Patients with prior surgery to the imaged knee were excluded. Patients were propensity score matched 1:1:1 by age, sex, and body mass index (BMI). Measures of patellar height, position, translocation, and trochlear dysplasia were collected from MRI performed prior to surgical intervention and compared between patients with patellofemoral OCD and PFI, and patellofemoral OCD and controls. Statistical analysis was performed in R Studio 4.2.1.
Results:
108 patients in each group (324 patients total, mean age 14.3 ± 2.3 years, mean BMI 22.8 ± 5.8, 154 [48%] female) met inclusion criteria. Caton-Deschamps index (CDI), patellar tilt, and tibial tuberosity-trochlear groove (TT-TG) distance were significantly higher in patients with OCD compared to controls, though were within normal range (CDI: controls 1.0 ± 0.16 vs OCD 1.1 ± 0.17, p<0.001; patellar tilt: control 10.0 ± 5.1 vs OCD 11.7 ± 6.0, p=0.03; TT-TG: control 12.9 ± 3.9 vs OCD 14.7 ± 4.2, p=0.002). All other measures of patellar/trochlear dysplasia were not significantly different between control and OCD patients. All measures of patellar and trochlear morphology differed significantly between OCD and PFI groups, and control and PFI groups. Logistic regression analysis indicated significantly increased risk of patellofemoral OCD in patients with elevated CDI (OR 1.30, p<0.001), patellar tilt (OR 1.41, p<0.001), and TT-TG (1.24, p<0.001) relative to controls.
Conclusion:
Patients with patellofemoral OCD have knee anatomy that appears to be distinct from both patients with PFI and controls. This altered morphology compared to healthy controls may lead to abnormal patellar tracking and increase the risk of developing patellofemoral OCD. Further research is needed to better characterize anatomic risk factors associated with the development of patellofemoral OCD.
