Patella Alta on X-Ray and MRI: Diagnostic Agreement and Clinical Correlations.

Background/

Objectives

Patella alta contributes to patellofemoral (PF) maltracking and anterior knee pain, but the agreement of patellar-height indices across imaging modalities and their clinical relevance remain uncertain. We aimed to estimate the prevalence of patella alta when using each index, quantify inter-method agreement, and quantify patella alta's associations with Hoffa's fat-pad impingement, patellar chondromalacia, and tibial tubercle-trochlear groove (TT-TG) distance.

Methods

This retrospective cross-sectional study included 250 patients aged 18-38 years who underwent both knee radiography and MRI. Patellar height was assessed using the Insall-Salvati (IS) and modified Insall-Salvati (mIS) ratios on radiographs and MRI, and the Patellotrochlear Index (PTI) on MRI. Associations with fat-pad impingement, chondromalacia grade, and TT-TG abnormality were analyzed using the chi-square or Fisher's exact test and as odds ratios (ORs) adjusted for age, sex, and imaged side; agreement was assessed with Cohen's kappa (κ).

Results

Patella alta prevalence was 13.2% using IS on radiographs and 10.8% using IS on MRI, but markedly lower with mIS (≤1.6%). Patella alta was more frequent in patients with Hoffa's fat-pad impingement (31.1% vs. 5.7% on IS radiography; adjusted OR 8.83, 95% CI 3.82-20.45) and advanced chondromalacia (p = 0.025). Agreement was moderate between IS radiography and IS MRI (κ = 0.47, 95% CI 0.33-0.61) and between IS MRI and the PTI (κ = 0.43).

Conclusions

The prevalence and clinical relevance of patella alta depend strongly on the measurement method. The IS ratio was the most consistent index, but radiography-MRI agreement was moderate rather than high, so the two are not interchangeable. The PTI adds information on trochlear engagement from the same MRI examination. These cross-sectional associations do not establish causation.