Less Invasive Rotational Acetabular Osteotomy for Hip Dysplasia
- First Online:
- Cite this article as:
- Maruyama, M., Wakabayashi, S. & Tensho, K. Clin Orthop Relat Res (2013) 471: 1263. doi:10.1007/s11999-012-2599-6
- 1.1k Downloads
Broad dissection with a long skin incision and detachment of the gluteus medius muscle performed for rotational acetabular osteotomy (RAO) can result in weakness in abduction strength of the hip. We use a surgical procedure for RAO that minimizes operative invasion of soft tissue and reduces incision length compared with conventional procedures.
We evaluated the clinical results of this less-invasive RAO comparing it with the more-invasive prior procedure with respect to improvement in clinical hip scores and radiographic coverage and overall hip survival after the procedure.
In this less-invasive exposure, the medial gluteus muscle is retracted to expose the ilium without detachment from the iliac crest. Similarly, the rectus femoris muscle tendon is retracted, not excised. The lateral part of the osteotomized ilium is cut to form the bone graft instead of harvesting it from the outer cortical bone of the ilium. Between 2000 and 2009, 62 patients (71 hips) underwent this procedure. Twenty-eight hips had early-stage osteoarthritis and 43 had advanced-stage osteoarthritis. Mean patient age was 40 years at the time of surgery. We evaluated improvement in hip scores (Merle d’Aubigné-Postel, Japanese Orthopaedic Association) and radiographic appearance (lateral center-edge angle, Sharp’s angle, acetabular head index [AHI]). Kaplan-Meier survivorship analysis was performed. Mean followup was 5 years (range, 2.0–10.4 years).
Clinical hip scores improved postoperatively. On average, lateral center-edge angle, Sharp’s angle, and AHI improved by 38°, 11°, and 42%, respectively. Predicted 10-year survival rates were 100% and 72% for hips with early- and advanced-stage osteoarthritis, respectively.
In hips with early-stage osteoarthritis treated by this less-invasive approach, no progression of osteoarthritis was documented and Trendelenburg gait was avoided. However, further investigation is necessary for hips with advanced-stage osteoarthritis.
Level of Evidence
Level IV, therapeutic study. See Guidelines for Authors for a complete description of levels of evidence.
Although minimally invasive surgery is commonly used in THAs [8, 19, 22], there are only a few reports regarding minimally invasive surgery for acetabular reorientation osteotomy [2, 20, 28, 29]. There are several reports regarding acetabular redirectional osteotomies for hip dysplasia, such as the triple osteotomy , dial osteotomy , double osteotomy , spherical osteotomy , rotational osteotomy [18, 26, 27], and periacetabular osteotomy , which are recognized standard surgical procedures for early stages of degenerative osteoarthritis. In adult hip disease, acetabular dysplasia can lead to early-onset degenerative joint disease [5, 30, 34]. Development of advanced osteoarthritis in patients with hip dysplasia has been reported to range from 20% to 50% [3, 4, 14]. Rotational acetabular osteotomy (RAO) initially was described by Ninomiya and Tagawa [18, 27], and now is recognized as an effective reconstructive surgical procedure for treatment of early- or advanced-stage osteoarthritis secondary to developmental acetabular dysplasia [36, 37].
However, this procedure as described [18, 26, 27] requires broad dissection of the ilium, ischium, and pubis and detachment of muscle insertions for adequate exposure and to allow for formation of the bone graft from the outer cortical bone of the ilium [1, 36, 37]. The detachment of the gluteus medius muscle might result in substantial weakness in the abduction strength of the hip. There have been few reports  evaluating postoperative abductor muscle function after conventional RAO [18, 36]. In addition, the intraarticular osteotomy used in the original RAO [18, 26, 27] left the teardrop (acetabular floor) in its original position and possibly resulted in deprivation of the osteotomized acetabulum from its blood supply, except for that coming from the capsule. The techniques, such as that of Ganz et al. , require extensive exposure and soft tissue release with a long skin incision of 25 cm or more [9, 18, 26, 37]. We believe these extensive exposures impair functional results and lead to cosmetically compromising scars . Teratani et al. [28, 29] reported a less-invasive periacetabular osteotomy with a skin incision length of 9 cm for all cases. Although this is an excellent procedure, it is highly technically demanding, especially in obese patients.
Comparison of conventional RAO with our surgical procedure
Length of skin incision (cm)
Number of incisions of the fascia lata
One incision or a y-shaped incision just below the skin incision
Transtrochanteric approach (trochanter osteotomy)
Detachment of the medial gluteal muscle from the ilium
Yes, partially, 7- to 10-cm width from the anterior superior iliac spine*
Detachment of the rectus femoris muscle
Division of the external short rotator muscles
None except for the piriformis
Thickness of the osteotomized acetabulum cephalad to the joint space (cm)
1.5*–2.5† (1* to 1.5† finger breadths)
2.5 (1.5 finger breadths)
Bone graft for gap between the osteotomized acetabulum and the ilium
Yes; one or two trapezoidal bone grafts from the outer table of the iliac wing
Yes; one trapezoidal bone graft from the lateral part of the osteotomized acetabulum
The purpose of our study was to describe this less-invasive surgical technique and report on the improvements in clinical hip scores, noting especially the presence or absence of a Trendelenburg gait. Our second aim was to assess the reconstruction radiographically, and third, to assess hip survivorship after the procedure. Finally, we compared this new approach with the more-invasive exposure using historical records and patients having the prior exposure on the contralateral side.
Patients and Methods
Patient demographics and results
Number of patients/hips
Female:male (number of patients)
Age at time of surgery (years)*
39.7 ± 10.3 (19–65)
157.2 ± 6.7 (140.9–169.5)
56.5 ± 10.0 (38.0–83.0)
22.9 ± 3.9 (16.3–31.8)
Osteoarthritis stage (number of hips)
1 + 2
5.3 ± 2.7 (2.0–10.4)
Warfarin was used routinely for deep venous thrombosis prophylaxis, and a few patients recognized as being at higher risk were administered low-molecular-weight heparin for 2 weeks.
Cast immobilization was not used, but postoperative traction was maintained for 10 days of bed rest to avoid excessive stress to the rotated acetabular fragment. Patients were allowed to use a wheelchair with touch-down weightbearing after passive and assisted active ROM exercises as instructed by our physiotherapists. Partial weightbearing ambulation with the aid of crutches was begun 5 to 6 weeks after surgery: 1/6 body weight on the affected leg for 6 to 8 weeks followed by an increasing weightbearing program supervised by the physiotherapists under the guidance of the operating surgeon (MM). Hardware, such as Kirschner wires and CCH screws, routinely were removed 7 weeks postoperatively.
Clinical followup was performed semiannually using the Merle d’Aubigné-Postel system  and the Japanese Orthopaedic Association (JOA) hip score  with evaluation of ambulation for the presence of a Trendelenburg gait by the operating surgeon (MM) and his colleagues. The hips were assessed radiographically to evaluate correction of deformity, measuring the lateral center-edge (CE) angle of Wiberg , Sharp’s angle , and acetabular head index (AHI) of Heyman and Herndon  preoperatively and postoperatively. One of us (MM) performed the radiographic measurements using a digital caliper (Mitsutoyo, Tokyo, Japan) with an accuracy of ± 0.02 mm. Healing of the osteotomy site and progression of osteoarthritis also were assessed on postoperative radiographs semiannually.
We used 2010 version SSRI® software (SSRI Inc, Tokyo, Japan) to perform the statistical analysis and Excel® (Microsoft Corp, Redmond, WA, USA) for descriptive statistics. Paired t-tests were used to compare radiographic parameters (lateral CE angle, Sharp’s angle, AHI) and clinical hip scores (Merle d’Aubigné-Postel, JOA) between preoperative and postoperative groups. We performed Kaplan-Meier survivorship analysis to estimate the probability of progression and change of the radiographic osteoarthritis stage. The association of preoperative joint disease with postoperative clinical scores less than those preoperatively and radiographic signs of progression of osteoarthritis as the end point was assessed with the log-rank test. The 95% CIs were calculated. Ten patients in this series with previous surgery on the contralateral hip by the conventional surgical procedure developed by Ninomiya and Tagawa [18, 26] or by the modified Ollier`s transtrochanteric approach [15, 16] with partial detachment of the gluteus medius muscle from the ilium were analyzed using Fisher’s exact test and the Mantel-Haenszel test. For all analyses, statistical significance was defined as p values less than 0.05.
The average operative scar lengths measured at latest followup were 17.3 ± 2.2 cm (range, 15.5–21 cm) for 33 hips operated on between August 2000 and August 2004 and 12.9 ± 1.4 cm (range, 10–15 cm) for 38 hips operated on between September 2004 and April 2009. The length of the skin incision was gradually shortened secondary to the surgeon becoming more proficient in the surgical procedure. There were no major neurovascular injuries or intraarticular penetration of the osteotome intraoperatively.
Clinical results after RAO
Merle d’Aubigné-Postel score (points)
11.1 ± 1.7 (6–14)
17.8 ± 0.2 (17–18)
10.6 ± 1.6 (7–13)
15.4 ± 2.7 (8–18)
JOA hip score (points)
58.4 ± 9.9 (30–78)
98.2 ± 1.8 (95–100)
56.0 ± 9.1 (31–74)
87.1 ± 14.6 (43–100)
Radiographic results after RAO
Lateral CE angle (°)
−1.3 ± 10.1 (−44 to 15)
36.5 ± 8.9 (15–52)
Sharp’s angle (°)
50.3 ± 4.3 (40–64)
39.4 ± 4.8 (26–50)
54.0 ± 10.7 (13–85)
95.7 ± 7.9 (81–113)
All 10 patients who had undergone previous surgery in the contralateral hip were more satisfied with the cosmetic appearance and scar length with the current procedure (mean length, 13.5 cm) than with the previous scar length (mean length, 26.5 cm). No Trendelenburg gait was observed in patients who underwent the current procedure but four patients (four hips) who had the previous technique had a Trendelenburg gait (Fisher’s exact test, p < 0.05; Mantel-Haenszel test, p < 0.03).
Broad dissection with a long skin incision and detachment of the gluteus medius muscle performed for RAO can result in weakness in the abduction strength of the hip. We aimed to reduce surgical invasion of conventional acetabular osteotomies by developing a modified approach and technique. The advantages of the new technique are as follows: (1) one approach with a relatively short skin incision is used, which provides enough exposure by osteotomizing the greater trochanter; (2) the muscle power of the hip abductor is preserved to prevent a postoperative Trendelenburg gait by retraction of the medial gluteus muscle without detachment from the ilium and iliac crest; (3) blood supply from the rectus femoris muscle to the osteotomized acetabulum is preserved as the rectus femoris tendon is not detached from the anterior inferior iliac spine, and blood supply to the capsule also is preserved as the osteotomy is extraarticular, including the acetabular floor (ie, teardrop in radiograph); and (4) if necessary, the lateral part of the osteotomized ilium (acetabulum) is cut in lunate (lateral view) and trapezoid (AP view) form for the bone graft instead of the outer cortical bone of the ilium to avoid excessive exposure. A fifth possible advantage of this procedure that was not evaluated in this study is that the shape of the pelvis, especially the pelvic ring, is unchanged, permitting normal vaginal delivery, a factor of importance, as most patients with acetabular dysplasia are adolescent and young women. By using this procedure, the patients in our series achieved good clinical hip scores and hip survivorship with reduction of postoperative Trendelenburg gait.
However, this study has several limitations including a relatively short followup, small numbers of subjects especially for comparison to conventional RAO, and potential observer bias.
The total length of the skin incision was 25 cm  for the bikini incision and conventional RAO. In the case of an ilioinguinal incision through a modified Smith-Petersen approach, such as a Bernese periacetabular osteotomy (Ganz), the skin incision length was at least 20 cm or more [7, 13, 21]. The advantages of trochanteric osteotomy are that it allows a visible continuous osteotomy line in an operative field, and distal transposition (ie, pulling down) of the greater trochanter is available to control tension of the hip abductor muscle. A disadvantage is that it takes one more procedure for reattachment of the greater trochanter. Our surgical procedure was not minimally invasive because muscle resection of the gluteus minimus and short rotator (piriformis) were inevitable, although these resections and reattachments do not influence abductor strength. As this is one limitation of our study, we called this procedure less invasive.
The predicted 10-year survival rate in this study was 72.1% for advanced-stage osteoarthritis hips, which is similar to rates of 72.2% after 10 years , 79% after 11 years , and 71% after 8 years  in other reports. In addition, the affected hips functioned well clinically without weakening the hip abductor muscles in almost all cases. Thus, it seems there is no disadvantage of this less-invasive procedure in patients with Stages 1 to 3 osteoarthritis. However, on the basis of our data, this type of osteotomy is not recommended for patients with more advanced-stage osteoarthritis (Stage 4) who had aggravation of joint incongruency, narrowing of the joint space less than 1 mm observed on AP radiographs of the patient in the standing position, and especially partial disappearance (Class D congruency according to Yasunaga et al. ) of the cartilage space in the hip abducted position.
Our approach requires a single skin incision of only 10- to 15-cm length, which provides satisfactory exposure by only retracting the gluteus medius with an osteotomized greater trochanter and an approach only from the outside of the ilium with a osteotome. Although this surgical technique might provide a benefit to the patients undergoing RAO, especially for cosmesis, which is important to younger female patients, further investigation is necessary to examine the benefit relative to long-term clinical and radiographic results.
Our surgical procedure focused on a short skin incision, retraction (not excision) of the gluteus medius muscle with an osteotomized greater trochanter, no excision of the rectus femoris muscle, and an approach only from the outside of the ilium with a osteotome. We had no patients with Trendelenburg gait postoperatively and good clinical results without progression of osteoarthritis in hips with early-stage ostearthritis. However, further investigation is necessary for hips with advanced-stage osteoarthritis.
We thank Judy R. Feinberg PhD, for excellent assistance and advice. We also thank Shizuo Murase MD, Ichiro Owan MD, and Naoyuki Katayama MD, for their clinical assistances.
This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.