Offset and anteversion reconstruction after cemented and uncemented total hip arthroplasty: an evaluation with the low-dose EOS system comparing two- and three-dimensional imaging.

PURPOSE: Accurate evaluation of femoral offset is difficult with conventional anteroposterior (AP) X-rays. The EOS imaging system is a system that makes the acquisition of simultaneous and orthogonal AP and lateral images of the patient in the standing position possible. These two-dimensional (2D) images are equivalent to standard plane X-rays. Three-dimensional (3D) reconstructions are obtained from these paired images according to a validated protocol. This prospective study explores the value of the EOS imaging system for comparing measurements of femoral offset from these 2D images and the 3D reconstructions.

METHODS: We included 110 patients with unilateral total hip arthroplasty (THA). The 2D offset was measured on the AP view with the same protocol as for standard X-rays. The 3D offset was calculated from the reconstructions based on the orthogonal AP and lateral views. Reproducibility and repeatability studies were conducted for each measurement. We compared the 2D and 3D offset for both hips (with and without THA).

RESULTS: For the global series (110 hips with and 110 without THA), 2D offset was 40 mm (SD 7.3; 7-57 mm). The standard deviation was 6.5 mm for repeatability and 7.5 mm for reproducibility. Three-dimensional offset was 43 mm (SD 6.6; 22-62 mm), with a standard deviation of 4.6 for repeatability and 5.5 for reproducibility. Two-dimensional offset for the hips without THA was 40 mm (SD 7.0; 26-56 mm), and 3D offset 43 mm (SD 6.6; 28-62 mm). For THA side, 2D offset was 41 mm (SD 8.2; 7-57 mm) and 3D offset 45 mm (SD 4.8; 22-61 mm). Comparison of the two protocols shows a significant difference between the 2D and 3D measurements, with the 3D offset having higher values. Comparison of the side with and without surgery for each case showed a 5-mm deficit for the offset in 35 % of the patients according to the 2D measurement but in only 26 % according to the 3D calculation.

CONCLUSIONS: This study points out the limitations of 2D measurements of femoral offset on standard plane X-rays. The reliability of the EOS 3D models has been previously demonstrated with CT scan reconstructions as a reference. The EOS imaging system could be an option for obtaining accurate and reliable offset measurements while significantly limiting the patient's exposure to radiation.