Patellofemoral morphology measurements and their associations with tibiofemoral osteoarthritis-related structural damage: exploratory analysis on the osteoarthritis initiative

Arya Haj-Mirzaian, Ali Guermazi, Farhad Pishgar, Frank W. Roemer, Christopher Sereni, Michael Hakky, Bashir Zikria, Shadpour Demehri

Research output: Contribution to journalArticlepeer-review


Objectives: Given the coexistence and possible interactions between patellofemoral and tibiofemoral compartments, roles of patellofemoral morphology measurements in tibiofemoral osteoarthritis (OA) have not been investigated extensively. We aimed to determine whether patellofemoral morphology is associated with the presence and longitudinal worsening of tibiofemoral OA in participants of the Osteoarthritis Initiative (OAI). Methods: Baseline knee MRIs of 600 participants were read by two independent blinded observers in consensus to determine patellofemoral morphology measurements including tibial tuberosity to trochlear groove (TT–TG) distance, trochlear groove depth (TGD), lateral patellar tilt (LPT), and Insall–Salvati ratio (ISR). Radiographic and MRI OA knee scoring (MOAKS) measurements were extracted from baseline and 2-year follow-up readings. Associations between baseline patellofemoral morphology metrics with radiographic medial tibiofemoral compartment (MTFC) joint space loss (> 0.7 mm, between baseline and 2nd–4th-year readings), and MRI-derived cartilage damage, bone marrow lesions (BMLs), and osteophytes (baseline to 2 years), were investigated using regression models adjusted for age, sex, body mass index, and knee alignment. P values were corrected using the Benjamini–Hochberg procedure. Results: Patellofemoral morphology measurements were not associated with longitudinal joint space loss in the MTFC or MOAKS determinants. Only TT–TG distance was associated with the baseline number of subregions with cartilage defects (OR (95% CI), 1.09 (1.04–1.14), corrected p value ≤ 0.01), BMLs (OR (95% CI), 1.1 (1.04–1.17), corrected p value = 0.01), and osteophytes (OR (95% CI), 1.09 (1.05–1.14), corrected p value ≤ 0.01) in the lateral tibiofemoral compartment (LTFC), and worsening of LTFC cartilage defects over 2 years (OR (95% CI), 1.09 (1.03–1.16), corrected p value = 0.02). Conclusions: Higher TT–TG distance was associated with concurrent MRI-derived OA-related structural damages and 2-year follow-up worsening only in LTFC. No associations were detected between patellofemoral morphology measurements and MTFC OA progression. Key Points: • Of all patellofemoral morphology measurements, the only lateralization of the tibial tubercle may be considered as a risk factor for lateral (not medial) tibiofemoral osteoarthritis worsening. • Patellofemoral morphology measurements of patella alta, trochlear dysplasia, patellar tilt, and lateralization of the tibial tubercle are not associated with radiographic and MRI-based medial tibiofemoral osteoarthritis worsening over 2 years. • Using longitudinal MRI data, each millimeter increase of TT–TG distance is associated with a 9% (95% confidence interval, 3–16%) increase in odds of longitudinal cartilage defects in the lateral tibiofemoral (but not medial) compartment over 2 years.

Original languageEnglish (US)
Pages (from-to)128-140
Number of pages13
JournalEuropean radiology
Issue number1
StatePublished - Jan 1 2020


  • Knee
  • Magnetic resonance imaging
  • Osteoarthritis

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging


Dive into the research topics of 'Patellofemoral morphology measurements and their associations with tibiofemoral osteoarthritis-related structural damage: exploratory analysis on the osteoarthritis initiative'. Together they form a unique fingerprint.

Cite this