For inclusion in this retrospective, cross-sectional study, we reviewed the surgical reports and preoperative radiographs of 149 consecutive hips of patients who received osteochondroplasty for femoral offset reduction either by surgical hip dislocation or via an anterior approach from November 2003 through January 2010. The indication for surgery was FAI in all cases. The diagnosis of FAI was based on patient history (especially groin pain during strenuous physical activities and prolonged sitting), clinical (especially a positive impingement test), and radiographic findings (offset reduction in cam-type and either acetabular retroversion or excessive global femoral head coverage in pincer-type FAI). An anterior approach was used if the indication was only to restore anterior femoral offset either alone or in combination with a partial anterior labral resection or débridement. All patients with marked deformity of the proximal femur necessitating wider exposure, suspected cartilage lesions, or planned rim trimming received surgical hip dislocation. Additional inclusion criteria for the study included a comprehensive description of the condition of the labrum in the surgical records and a complete set of preoperative radiographs, consisting of an AP view of the pelvis and an axial Lauenstein view. The latter is the standard axial view at our institution and is about comparable to the 45° Dunn view, which was the most sensitive axial projection for detection of offset reduction in the study by Meyer et al. [20
]. The Lauenstein view has been used at other institutions for similar purposes [17
]. Patients with malrotated (foramen obturatum index < 0.56 or > 1.8) [27
] or malinclined pelvic radiographs (distance between tip of coccygeum and symphysis pubis < 0 or > 2 cm) and those who have had previous surgical treatment of the affected hip during adulthood were excluded from the study. We excluded 12 patients with previous operations (including pelvic osteotomies in seven patients and intertrochanteric osteotomies in two) and 14 patients with inadequate preoperative radiographs. Thus, 123 hips in 116 patients, 44 women and 72 men, with a mean age of 37.4 ± 9.5 years (range, 18–57 years), were included in the analysis. There were 51 left and 72 right hips affected. Twenty-one hips had cam-type, 17 hips had pincer-type, and 85 hips had mixed-type FAI. No patient had a history of major trauma preceding clinical symptoms. Ninety-two hips were operated on via surgical dislocation and 31 via an anterior approach.
We screened the surgical reports forms for the status of the acetabular labrum. A hook probe was used routinely intraoperatively to identify tears in the substance of the labrum or at the junction between the labrum and cartilage. Using an anterior approach, the undersurface of the labrum could be probed by rotating the hip with varying degrees of flexion in slight abduction. This sometimes was possible only after offset reduction. The presence, size, and location of labral tears were documented in the surgical reports. In the first step of the statistical analysis, the labrum was classified to be intact or flawed (torn, severely degenerated, or frayed). The classification scheme of Beck et al. [1
] also was used and modified for statistical analysis: an intact labrum received Grade 0, a labrum with degeneration or ossification received Grade 1, and a labrum with a full-thickness tear or detached from the acetabulum was labeled Grade 2 (Table ). Acetabular chondromalacia was graded according to Outerbridge [22
]. As the anterior approach does not allow inspection of the acetabular cartilage, patients treated by the anterior approach were not included in the subanalysis on the coincidence of cartilage and labral lesions.
Modification of the classification scheme of Beck et al.
The radiographic measurements were made by an orthopaedic surgeon trained in hip-preserving surgery (TK). From the AP pelvis, the acetabular index of the weightbearing zone [2
], center-edge angle [30
], inclination of the acetabulum [26
], lateral head extrusion index [10
], neck-shaft angle, asphericity of the femoral head [4
], superior alpha angle [21
], superior offset, and superior offset ratios [5
] were acquired using commercially available software (DiagnostiX-32, Version 3.8.6, 2003; Gemed, Ulm, Germany). Asphericity was determined to be present when the junction of the femoral head and neck lay outside a circle matching the surface of the femoral head [4
]. The alpha angle according to Nötzli et al. [21
] is the angle between a line connecting the narrowest point of the femoral neck to the center of the femoral head and a line connecting the latter to the point where the surface of the femoral head leaves a circle at the head-neck junction. Femoral offset was measured according to the original definition of Eijer et al. [5
] as the distance between a line parallel to the femoral neck axis at the cortex of the femoral neck and a parallel tangent to the femoral head. The offset ratio was defined as the ratio between the femoral offset and the diameter of the femoral head [5
]. Retroversion of the acetabulum was assumed when the anterior and posterior acetabular rim crossed (crossover sign) [14
] or the ischial spine projected into the pelvic cavity (ischial spine sign) [12
]. The posterior wall sign was assumed to be present when the center of the femoral head lay lateral to the posterior acetabular rim [14
On the axial view, asphericity of the femoral head, anterior alpha angle, anterior offset, and anterior offset ratio were determined in a similar fashion. The extent of radiographic osteoarthritis was assessed according to Tönnis [27
] and Kellgren and Lawrence [13
]. In a previous study, the 95% confidence interval of radiographic measurements for the same observer and the identical measuring technique had been determined to be −0.5° and +0.9° for acetabular index of the weightbearing zone, −1.9° and +0.9° for center-edge angle, −0.5° and +1.5° for acetabular inclination, −0.02 and +0.06 for lateral head extrusion index, −1.0° and +1.3° for neck-shaft angle, −1.7° and +1.5° for superior and −1.5° and +1.9° for anterior alpha angle, −1.2 and +1.1 mm for superior and −0.7 and +1.3 mm for anterior offset, and −0.03 and +0.03 for superior and −0.03 and +0.04 for anterior offset ratio. We determined Kappa values to be 1.000 for the assessment of asphericity, 1.000 for the Tönnis classification, and 0.750 for the Kellgren and Lawrence classification (Kappe T, Kocak T, Fraitzl CR, Reichel H. Radiological risk factors for cartilage lesions in femoroacetabular impingement. Poster presentation P028, San Diego, CA, AAOS Annual Meeting February 15-19, 2011.).
Radiographic measurements in patient groups with and without labral alterations or tears were compared using one-way ANOVA for multiple groups, Student’s t test (metrical data), and the chi square test (nominal data). Established cutoff values were used to delineate pathologic from nonpathologic groups: less than 0° and greater than 10° for acetabular index of the weightbearing zone, greater than 39° and less than 25° for center-edge angle, less than 0.10 and greater than 0.25 for lateral head extrusion index, less than 126° and greater than 139° for neck-shaft angle, greater than 50° for alpha angle, less than 9 mm for offset, and less than 0.17 for offset ratio [9
software (Version 17.0; SPSS Inc, Chicago, IL, USA) was used for statistical analysis and significance assumed for a p value less than 0.05. In an a priori power analysis (G*Power, Version 3.1.2; Universität Kiel Dusseldorf, Germany), the necessary sample size for an effect size of 0.5, an alpha of 0.05, and a beta of 0.2 was determined to be 51.