¹Department of Orthopaedic Surgery, Inselspital, University of Bern, Bern, Schweiz
²University Institute of Diagnostic, Interventional and Paediatric Radiology, Inselspital, University of Bern, Bern, Schweiz
³Radiology Department, Balgrist University Hospital, Zürich, Schweiz

Text

Objectives and questions: Femoral torsion affects hip biomechanics and is an essential component in diagnosing mechanical hip pain, alongside assessing proximal femur deformities and acetabular version. Accurate measurement of femoral torsion is critical, with both CT and MRI commonly used as imaging modalities. High agreement, reliability, and reproducibility have been observed for femoral torsion measurements between CT and MRI. However, variations in anatomical landmarks across different measurement methods can result in discrepancies of up to 20° in reported torsion values. Furthermore, no studies have established normal reference standards for the most commonly used measurement techniques.

This study compares femoral torsion measurements using four MRI-based methods in asymptomatic volunteers and corresponding CT-based methods in patients with hip pain (Murphy et al., Tomczak et al., Reikerås et al., and Lee et al.).

Material and methods: This retrospective analysis included 47 asymptomatic volunteers (47 hips, mean age 30±5 years, 49% female) who underwent 3T MRI and 46 patients (52 hips, mean age 28±9 years, 59% female) from the clinic's database who underwent CT for routine hip pain evaluation. Patients had no childhood, inflammatory, or post-traumatic deformities. Volunteers were included if they had no hip-related symptoms, relevant history, or positive impingement test. Femoral torsion was measured using four methods, varying by proximal femoral neck landmark—from the most proximal (Lee) to the most distal (Murphy at the lesser trochanter). Femoral head center and distal reference points were consistent across methods. Statistical analysis used t-tests.

Results: Femoral torsion values increased with more distal landmarks in both groups, ranging from the lowest values using the Lee method (12.8 ± 10.8° in volunteers, 11.3 ± 10.8° in patients) to the highest with the Murphy method (32.9 ± 13.2° in volunteers, 26.0 ± 12.6° in patients). Torsion measurements were higher in asymptomatic volunteers compared to patients for the Murphy (32.9±13.2° vs. 25.9±12.6°, p=0.009) and Tomczak (29.3±10.8° vs. 24.0±12.4°, p=0.025) methods. No significant differences were found between volunteers and patients for the Reikerås (18.0±11.4° vs. 14.6±10.5°) and Lee (12.8±10.8° vs. 12.0±10.8°) methods.

Discussion and conclusions: More distal measurement methods (Murphy and Tomczak) appear more sensitive in detecting differences in femoral torsion between asymptomatic individuals and patients, suggesting these methods may more accurately capture clinically relevant variations linked to torsional deformities.