Femoroacetabular impingement (FAI) and developmental dysplasia of the hip (DDH) are major etiological factors in the development of osteoarthritis (OA) of the hip.

The pathomechanism of FAI is a dynamic condition caused by an altered osseous morphology of the acetabulum or femoral head-neck junction [1] -[3] . Two types of FAI can be distinguished: cam impingement and pincer impingement. Pincer impingement is caused by over coverage of the acetabulum relative to the femoral head, and cam impingement is caused by extra bone formation in the anterolateral head-neck junction.

In recent epidemiological studies, the significance of acetabular retroversion in DDH associated with FAI has been addressed [4] - [7] . FAI is thought to be associated with the pathology of DDH, but not much has been published about the morphological characteristics of the acetabulum and femoral head-neck junction in patients with hip dysplasia. This concept is important, because acetabular malpositioning after osteotomy may lead to an iatrogenic pincer type of impingement, especially in cases with lateral over coverage or retroversion [8] - [10] . In addition, an aspherical femoral head in DDH [11] will increase the risk of secondary cam-type FAI that is reportedly symptomatic in 22/46 (47.8%) hips after reorientation [7] . These mechanisms seem to be involved in secondary lesions of the acetabular labrum tear and the articular cartilage damage of the acetabulum, with subsequent development of OA. Several studies suggest that even mild cam-type deformities can significantly lead to secondary OA of the hip [1] .

The purpose of this study was to evaluate the preoperative radiographs with cases of DDH leading to rotational acetabular osteotomy (RAO) or curved periacetabular osteotomy (CPO) [12] retrospectively, and examine the frequency of FAI related bone morphology in the acetabulum and femoral head-neck junction.

This study design was approved by Institutional Review Board of Hyogo College of Medicine.

We performed a retrospective examination of pre- and postoperative radiographs on 24 hips in 23 patients, which included 22 females and 1 male, who were underwent acetabular osteotomy. We have improved and changed surgical procedure since 2009. Therefore, osteotomy was performed on 10 hips with RAO and 14 hips with CPO between December 2004 and December 2012. Osteotomy was performed with the RAO from December 2004 to February 2009, and with the CPO from February 2009 to December 2012. The mean age of these patients upon presentation was 30.8 years (ranged 19 to 44 years). Minimum follow up periods are 2 years after surgery. The Tönnis classification was used to grade the extent of OA [13] [14] . Six hips had grade 0, and eighteen hips had grade 1 OA according to the Tönnis classification system. We excluded patients with moderate and severe hip osteoarthritis (Tönnis grade 2 and 3) and major femoral head deformities such as typical capital drop deformity, Perthes’ disease, or slipped capital femoral epiphysis (SCFE). All of the patients including in this study were diagnosed with dysplasia. The surgical indications for RAO and CPO included symptomatic acetabular dysplasia under the age of 45˚, a lateral center-edge (CE) angle of less than 20˚ and a Sharp angle of greater than 45˚ on anteroposterior (AP) radiographs (Figure 1), and the improvement of joint congruency on an AP radiograph in the abducted position. The radiological assessment of potential cam-type deformity was based on the measurement of the alpha angle on the cross-table lateral view. It was measured by first drawing the best fitting circle around the femoral head, then a line through the center of the neck and the center of the head. From the center of the femoral head a second line was drawn to the point where the superior surface of the head-neck junction first departs from the circle. The angle between these two lines is the alpha-angle (Figure 2). Several studies have defined the upper limit of normal for the alpha angle as being 50.5˚ [1] [15] , and we applied the cut-off angle of more than 50.5˚ to define cam impingement deformity [1] . On the AP pelvic radiograph, Sharp angle, CE angle, crossover sign, and posterior wall sign [16] (Figure 3) were evaluated. Acetabular retroversion was defined as the presence of a crossover sign [16] - [18] .

All of these signs are sensitive to the position of the pelvis relative to the plane of the X-ray beam and an as

sessment of this was undertaken. Rotation was checked by confirming the alignment of the coccyx and symphysis pubis. The extent of the pelvic inclination was judged according to a previously described method [19] . The distance between the pubic symphysis and sacrococcygeal joint was measured on each standard AP radiograph for comparison with the reported control values of 25 - 40 mm for men and 40 - 55 mm for women [19] . All of the radiographs were assessed by a single reviewer (R.K). Patients whose preoperative radiograph did not fulfill these criteria were excluded from this study.

The results of the preoperative measurements were as follows: Sharp angle 49.5˚ ± 2.5˚, CE angle 9.5˚ ± 7.2˚, and alpha-angle 67.6˚ ± 18.9˚ (Table 1, Table 2). Crossover signs were revealed in 7 hips (29.2%) and posterior wall signs in 16 hips (66.7%). Cam-type deformities with an alpha angle of ≥50.5˚ were observed in 19 hips (79.2%) (Table 3). Of these 19 hips with cam-type deformities of ≥50.5˚, 13 hips (54.2%) had an alpha angle of ≥70˚ and 6 hips (25.0%) had an alpha angle of ≥80˚ (Figure 4). Additionally, combined deformities both acetabulum and femoral head-neck junction were observed in 14 hips (58.3%), which was revealed either positive cross over sign or posterior wall sign in acetabulum and an alpha angle of ≥50.5˚ in femur. DDH related deformity were improved after surgery with Sharp angle 37.1˚ ± 4.9˚ and CE angle 32.9˚ ± 9.9˚ in postoperative

(+) = positive, (−) = negative.

radiograph. As determined using the Tönnis scale, no progression of osteoarthritis was found in 16 of the 24 hips, there was a one-grade progression in 8 hips. Among, 6 of 8 hips (75%) with progression of osteoarthritis were observed either positive cross over sign or posterior wall sign and an alpha angle of ≥50.5˚ (Table 4).

The prevalence of acetabular retroversion in the normal population as well as patients with hip disorders has been previously investigated [4] [5] . Fujii et al. [4] reported acetabular retroversion in 17 of the 96 hips (18%) of patients with DDH. Also, Ezoe et al. [5] reported the prevalence in 18% (13 of 74 hips) of DDH patients as compared to 6% (7 of 112 hips) in normal subjects. Compared with previous literatures, the background of the patients in this study could be possible severe DDH with Sharp angle 49.5˚ ± 2.5˚ and CE angle 9.5˚ ± 7.2˚, which need rotational osteotomy. The incidence of the preoperative retroversion in this study was relatively high (29.2%). Additionally, 18 hips (75%) were defined as acetabular side deformity under the hypothetical situation of either positive cross over sign or posterior wall sign. Regarding the surgical treatment for DDH, various forms of periacetabular osteotomy have been proposed and reported with their surgical outcome [12] [20] . Although these procedures can effectively improve lateral coverage of the femoral head, inadvertent post-operative acetabular retroversion has also been reported [6] [10] [21] . Myers et al. [10] reported five cases presenting anterior FAI caused by acetabular retroversion following the Bernese periacetabular osteotomy. Xie et al. [6] evaluated patients with DDH who underwent CPO and found acetabular retroversion in 16 (15%) of 106 hips preoperatively, while the postoperative evaluations showed 66 (62%) hips with acetabular retroversion. The evaluation of acetabular version revealed that 50 of 106 hips (47%) with preoperative anteversion were overcorrected to retroversion. Additionally, 14 hips had progressed arthrosis, and 10 of the 14 hips (71%) with progression of arthrosis showed positive crossover sign. Furthermore, Kiyama et al. [21] reported that 5 of 24 hips (24%) with acetabular retroversion following CPO were complicated with post-operative progression of osteoarthritis. On the other hand, Yasunaga et al. [7] reported that the rate of cases with a positive crossover sign showed an increase after RAO from 7.8% before surgery to 42.6% after surgery. Despite this, most of the cases with a positive crossover sign also showed a positive impingement sign; however, a positive cross over sign was not involved with the progression of arthrosis. Additionally, preoperative CT was necessary for the preoperative planning for osteotomy, we also examined the Preoperative CT for all patients including in the present study. However, no papers have been described the definition to quantify the retroversion and Cam type deformity by CT evaluation. Accordingly, we only evaluated the each parameter in the plain radiograph in the present study.

In addition, a high incidence of an impingement sign after PAO has been reported (24% - 47.8%) [8] [22] [23] (Table 5), however, the literatures have not mentioned an association with progression of OA, impingement sign, alpha angle, and crossover sign. Regarding cam-type deformity, Hack et al. [24] reported that deformity with an alpha angle of ≥50˚ was noted in 24.7% and 5.4% of the hips of healthy male and female volunteers, respectively. In Japanese patients, Mori et al. [25] reported the prevalence of cam type deformity (alpha angle ≥50.5˚) with Tönnis grade 0 and 1. They reported that 29 of 202 hips (14.1%) had cam type deformity. In our present study, the incidence of cam-type deformity was relatively high (19 of 24 hips 79.2%). In particular, the distribution of the alpha angle showed angles of ≥80˚ were present in 6 (25.0%) of 19 hips with an alpha angle of ≥50.5˚ in DDH patients.

Regarding the relationship between cam-type deformity and progression of OA, Agricola et al. [26] reported that the odds ratio of progression to end stage OA within 5 years with an alpha angle of ≥60˚ and ≥83˚ was 3.67 and 9.66, respectively, and concluded that alpha angle was strongly related to the progression of OA. Considering this report, correcting DDH without treatment of cam-type deformity may lead to be could be the risk of

progression of OA. In a recent report, Christoph et al. [27] evaluated proper acetabular reorientation and the creation of a spherical femoral head to improve long-term survivorship and decelerate OA progression in patients with DDH. In our present study, the presence of cam-type deformity in patients who underwent RAO or CPO was relatively high in preoperative radiographs. In addition, combined FAI related bone morphology in the acetabulum and femoral head-neck junction were showing either positive cross over sign or posterior wall sign and an alpha angle of ≥50.5˚ in 14 hips (58.3%), among 6 of these 14 hips (42.8%) were progressed OA grade. The morphologic features of the femoral head may lead to progression of OA without optimal acetabular orientation, which is meant to provide the best reasonable orientation given the anatomic constraints and corrected spherical femoral head. Currently, to prevent secondary FAI, we have been performing endoscopic osteochondroplasty before CPO for the cases with an alpha angle of ≥50.5˚ in preoperative radiograph [28] .

This study has several limitations. First, this retrospective study included a small sample size and lacked a control group with normal hips. Second, intraobserver reliability was not reported. Previous reports showed a poor correlation between the intraobserver reporting of the crossover sign and alpha-angle [29] . Third, our study did not evaluate pre- and postoperative physical examination. Finally, we could not evaluate the post-operative radiological result with long-term follow up, our minimum follow up periods are 2 years after surgery. In a future study we need to evaluate the relationship between the physical examination and radiological findings, and a long term follow up about progression of OA is also necessary.

We evaluated the morphological variables in patients with hip dysplasia who underwent rotational acetabular osteotomy. The results of this study support the hypothesis that there are differences in the morphology of the proximal femur and acetabulum in patients with hip dysplasia who underwent RAO and CPO compared to normal hips. Especially, the presence and severity of cam-type deformity were relatively higher in preoperative radiographs than other previous reports. There is a possibility of secondary FAI due to excessive forward coverage of the rotated/moved bone fragments and cam-type deformity after RAO and CPO, which is involved with the progression of OA. Therefore, secondary FAI should be prevented by adding to surgical methods with careful consideration of these findings.

The authors claim that there exist no conflicts of interest.