Surgical treatment for pincer femoroacetabular impingement (FAI) of the hip remains controversial, between trimming the prominent acetabular rim and reverse periacetabular osteotomy (PAO) that reorients the acetabulum. However, rim trimming may decrease articular surface size to a critical threshold where increased joint contact forces lead to joint degeneration. Therefore, knowledge of how much acetabular articular cartilage is available for resection is important when evaluating between the two surgical options. In addition, it remains unclear whether the acetabulum rim in pincer FAI is a prominent rim because of increased cartilage size or increased fossa size. We used reformatted MR and CT data to establish linear length dimensions of the lunate cartilage and cotyloid fossa in normal, dysplastic, and deep acetabula. We reviewed the last 200 hips undergoing PAO, reverse PAO, and surgical dislocation for acetabular rim trimming at one institution. We compared MR images of symptomatic hips with acetabular dysplasia (20 hips), pincer FAI (29 hips), and CT scans of asymptomatic hips from patients who underwent CT scans for reasons other than hip pain (20 hips). These hips were chosen sequentially from the underlying pool of 200 potential subjects to identify the first 10 male and the first 10 female hips in each group that met inclusion criteria. As a result of low numbers, we included all hips that had undergone reverse PAO and met inclusion criteria. Cartilage width was measured medially from the cotyloid fossa to the lateral labrochondral junction. Cotyloid fossa linear height was measured from superior to inferior and cotyloid fossa width was measured from anterior to posterior. Superior lunate cartilage width (SLCW) and cotyloid fossa height (CFH) were measured on MR and CT oblique coronal reformats; anterior lunate cartilage width (ALCW), posterior lunate cartilage width (PLCW), and cotyloid fossa width (CFW) were measured on MR and CT oblique axial reformats. Cohorts were compared using multivariate analysis of variance with Bonferroni's adjustment for multiple comparisons. Compared with control acetabula, dysplastic acetabula had smaller SLCW (2.08 ± 0.29 mm versus 2.63 ± 0.42 mm, mean difference = -0.55 mm; 95% confidence interval [CI] = -0.83 to -0.27; p < 0.01), ALCW (1.20 ± 0.34 mm versus 1.64 ± 0.21 mm, mean difference = -0.44 mm; 95% CI = -0.70 to -0.18; p = 0.00), CFH (2.84 ± 0.37 mm versus 3.42 ± 0.57 mm, mean difference = -0.59 mm; 95% CI = -0.96 to -0.21; p < 0.01), and CFW (1.98 ± 0.50 mm versus 2.77 ± 0.33 mm, mean difference = -0.80 mm; 95% CI = -1.16 to -0.42; p < 0.0001). Based on the results, we identified two subtypes of deep acetabula. Compared with controls, deep subtype 1 had normal CFH and CFW but increased ALCW (2.09 ± 0.42 mm versus 1.64 ± 0.21 mm; p < 0.001) and PLCW (2.32 ± 0.36 mm versus 2.00 ± 0.32 mm; p = 0.04). Compared with controls, deep subtype 2 had increased CFH (4.37 ± 0.51 mm versus 3.42 ± 0.57 mm; p < 0.01) and CFW (2.76 ± 0.54 mm versus 2.77 ± 0.33 mm; p = 1.0) but smaller SCLW (2.12 ± 0.40 mm versus 2.63 ± 0.42 mm; p < 0.01). Deep acetabula have two distinct morphologies: subtype 1 with increased anterior and posterior cartilage lengths and subtype 2 with a larger fossa in height and width and smaller superior cartilage length. In patients with deep subtype 1 hips that have increased anterior and posterior cartilage widths, rim trimming to create an articular surface of normal size may be reasonable. However, for patients with deep subtype 2 hips that have large fossas but do not have increased cartilage widths, we propose that a reverse PAO that reorients yet preserves the size of the articular surface may be more promising. However, these theories will need to be validated in well-controlled clinical studies.
Read full abstract