Standardized magnetic resonance image‐based assessment to define functional patella alta relative to both tibia and trochlea: A cross‐sectional comparative study

PURPOSE: Determine whether patellar height differs significantly between knees with objective patellar instability (OPI) and controls, and to develop a magnetic resonance imaging-based (MRI-based) quantitative method for classifying 'functional patella alta' relative to both tibia and trochlea. METHODS: This cross-sectional study included all records of adults who underwent an MRI of their knee between 2019 and 2022 at the senior authors' centre. The OPI group (n = 127) included patients with >2 documented episodes of lateral patellar dislocation and no previous knee surgery, and the control group (n = 97) included patients with isolated meniscal tears and no history of patellofemoral disorders or knee surgery. Four readers independently measured the MRI patellar height index (PHI), the sagittal patellar engagement (SPE) index and the patellar tendon length (PTL). The control group's interquartile ranges of PHI (75th percentile) and SPE index (25th percentile) were used to define 'patella norma' (PHI ≤ 1.16 and SPE index ≥ 0.38) and 'functional patella alta' (PHI > 1.16 and SPE index < 0.38). Multivariable logistic regression analyses assessed the associations between patellar height and trochlear dysplasia with knees exhibiting OPI. RESULTS: Comparison between 'patella norma' and 'functional patella alta' revealed statistically significant differences for PHI (mean difference [MD], -0.23; p < 0.001), SPE index (MD, 0.28; p < 0.001) and PTL (MD, -7; p < 0.001). The prevalence of 'functional patella alta' was 24% in OPI knees (30 of 127) and 6% in control knees (6 of 97). Multivariable logistic regression revealed that 'functional patella alta' and trochlear dysplasia were independently associated with OPI. CONCLUSION: Standardized MRI-based assessment of patella alta relative to both tibia and trochlea is reliable using thresholds of PHI (>1.16) and SPE index (<0.38). These thresholds, specifically developed for MRI, should improve the conventional assessment using the Caton-Deschamps index, originally developed for true lateral radiographs. LEVEL OF EVIDENCE: Level III.