The study protocol was approved by the institutional review board of our hospital, and informed consents were obtained from all patients. Between 2004 and 2008, the THA was performed in 52 patients (58 hips) with DDH. According to Crowe's classification,11) 21 hips (17 patients) with severe DDH (15 hips of Crowe type III and 6 hips of type IV) were retrospectively identified and reviewed. There were 14 women and 3 men with the mean age of 52.3 years (range, 34 to 72 years). All patients were followed for an average of 52 months (range, 38 to 72 months). Thirteen patients had unilateral involvements and 4 patients had bilateral involvements. No patient had a history of prior surgeries for DDH.

All operations were performed by a single senior surgeon using a modified Hardinge's lateral approach in the lateral decubitus position. Two-stage operations were planned in 6 irreducible hips with expected lengthening of the affected limb after THA of > 2.5 cm or with flexion contracture of > 30 degrees. In the 6 staged operations, surgical hip liberalization with abductor slide was performed through two skin incisions during the first stage (Fig. 1). Through a curvilinear incision over the iliac crest, abductors were detached from the ilium by subperiosteal dissections. This dissection was then continued inferiorly to the acetabular rim and posteriorly to the sciatic notch and the posterior superior iliac spine. The entire abductor origin was then mobilized from the pelvis without any tensions. After the femoral head had been resected through a modified Hardinge's lateral approach, the tensor fascia lata, sartorius muscle, iliopsoas tendon, and rectus femoris were sequentially released. Adductor tenotomy was additionally performed in cases with adductor contractures. After resecting the elongated hypertrophic joint capsule, the true acetabulum was identified and prepared using a small reamer. The trial acetabular component was then placed in the acetabulum to evaluate the bone stock. The resected femoral head was stored at our institutional bone bank for possible uses as a bulk autogenous bone graft to augment the acetabulum in the event of inadequate acetabular bone coverage for cup fixation. After the first operations, gradual skeletal traction was applied to the operated limb by placing a metal pin through the proximal tibia for 2 weeks. Traction weight was gradually increased to a maximum of one fourth of body weight, but reduced or removed when a patient complained of symptoms suspicious of neuralgia. A portable radiograph was taken to confirm that the greater trochanter tip retracted inferiorly to the center of the true acetabulum. During second operations, one hip was successfully reduced without femoral shortening, but 5 hips were difficult to reduce with trial components, and therefore, subtrochanteric shortening osteotomies were performed to prevent neurologic traction injuries. After subtrochanteric osteotomy, the onlay cortical bone graft with resected bone was affixed to the osteotomy site with cerclage cable (Fig. 2). The remaining 15 hips were treated by one-stage THA with soft-tissue release alone (Fig. 3). During the operative procedures, we did not explore nor palpate the sciatic nerve to evaluate tensions.

The acetabular cup was placed in the true acetabulum in all hips. Secure fixation of the cup was achieved in 11 of the 21 hips without bone grafting, while a bulk autogenous or allogenous femoral head bone graft was augmented to the acetabulum in 4 hips with unreliable fixation and insufficient bone coverage. Augmentation of the superolateral rim with cancellous bone graft was performed to increase the bone stock for future revisions in 6 hips with reliable fixation, despite insufficient bone coverage.

The cementless Duraloc acetabular component (DePuy, Warsaw, IN, USA) and Biolox forte (CeramTec, Plochingen, Germany) alumina-on-alumina bearing system was used in all hips. Femoral component fixation was uncemented in all hips. A straight, tapered, titanium Summit stem (DePuy) was used in 14 of the 21 hips, and the remaining 7 femoral components were S-ROM modular stems (DePuy). Femoral head size was 28 mm in all hips. Acetabular cups in the outer diameter ranged from 46 to 54 mm.

All patients were mobilized on the third day after THA. After one-stage THA, partial weight bearing with crutches was allowed from 5 days to 6 weeks, postoperatively. Thereafter, weight bearing was progressively increased. An abductor brace was applied to patients that underwent two-stage THA, and partial weight bearing with crutches was allowed from 7 days to 6 weeks. Full weight bearing was started based on radiographically evident osteotomy site consolidation.

In unilateral DDH cases, leg length discrepancy (LLD) was defined as the difference between the distances from the interteardrop line and the greater trochanter tip in affected and contralateral hips on anteroposterior hip radiographs, and was measured before and after surgery (Fig. 4). Clinical assessments were performed using Harris hip scores. The level of activities of the patients was assessed using the University of California, Los Angeles (UCLA) activity score at the latest follow-up.

All patients were reviewed radiologically at 6 weeks, 3 months and 6 months after surgery, and annually thereafter, and the presence of radiolucent lines around the acetabular component in the three zones defined by DeLee and Charnley16) was noted. Mode of fixation of the femoral component was classified as bone ingrowth, stable fibrous fixation, or unstable, as described by Engh et al.17) Osteolytic lesions were documented and classified by size and location.11) Serial radiographs were assessed to evaluate bony union at femoral osteotomy sites and at acetabular bone grafted sites. Data processing and statistical analyses were performed using SPSS ver. 17.0 (SPSS, Chicago, IL, USA). The Wilcoxon signed rank-sum test was used to compare preoperative and postoperative Harris hip scores. A p-value below 0.05 was considered significant.