Pre-operative templating in THA. Part I: a classification of architectural hip deformities

While numerous classifications of hip arthritis have been proposed, none considered the magnitude and direction of femoral head translation relative to the native acetabulum. A more precise classification of architectural hip deformities is necessary to improve preoperative templating and anticipate surgical challenges of total hip arthroplasty (THA). The purpose of the present study was to introduce a classification system to distinguish different types of architectural hip deformities, based on femoral head translation patterns, and to evaluate its repeatability using plain radiographs (qualitative) and Computed Tomography (CT) measurements (quantitative).

We studied pre-operative frontal and lateral hip radiographs and CT scans of 191 hips (184 patients) that received primary THA. The distance between the femoral head center (FC) and the acetabular center (AC) was measured, as well as femoral offset, acetabular offset, head center height, acetabular floor distance and femoral neck angle. The hips were classified qualitatively using frontal plain radiographs, and then quantitatively using CT scans (with an arbitrary threshold of 3 mm as Centered, Medialized, Lateralized, Proximalized or Proximo-lateralized. The agreement between qualitative and quantitative classification methods was compared for applying the same classification.

Qualitative classification identified 120 centered (63%), 8 medialized (4%), 49 lateralized (26%), 3 proximalized (2%), and 11 proximo-lateralized (6%) hips, while quantitative classification identified 116 centered (61%), 8 medialized (4%), 51 lateralized (27%), 5 proximalized (3%), and 11 proximo-lateralized (6%) hips. The agreement between the two methods was excellent (0.94; CI 0.90-0.98). Medialization reached 9.7 mm, while lateralization reached 10.9 mm, and proximalization reached 8.5 mm. Proximalized and proximo-lateralized hips had more valgus necks, while medialized hips had more varus necks (p = 0.003).

The classification system enabled repeatable distinction of 5 types of architectural hip deformities. The excellent agreement between quantitative and qualitative methods suggests that plain radiographs are sufficient to classify architectural hip deformities.