Study design
This study design was approved by the Institutional Review Board of Guangdong Provincial People’s Hospital and was performed in strict accordance with the ethical standards stipulated in the 1964 Declaration of Helsinki and its later amendments. Signed informed consent for participation was obtained from all study patients. The trial protocol was submitted to ClinicalTrials.gov, and the trial registration number is NCT01613612.
Based on the results of Gangi et al[14], we determined that at least 20 hips per group were necessary to detect a difference by using the Wilcoxon-Mann-Whitney test with a bilateral α of 0.017 and a power of 80%. We included a minimum of 26 hips in each group to compensate for the patients lost to follow-up.
Patients
We recruited eligible patients from those admitted to the Center of Orthopedic Surgery from June 2009 to October 2010. Plain radiographs of the bilateral hips in the anteroposterior and frog-leg lateral positions as well as magnetic resonance imaging (MRI) were performed in all patients. We confirmed the diagnosis of ANFH based on the clinical history and the radiographic lesions in the femoral head.
The inclusion criteria were as follows: the subjects (1) were 18 to 55 years old; (2) had notable hip pain; (3) had normal, minor, or mixed osteopenia or had crescent signs detected on plain radiographs; and (4) stopped steroid treatment for at least 6 months.
The exclusion criteria were as follows: (1) < 18 or > 55 years old; (2) terminal stage of ANFH with the presence of secondary osteoarthritic changes such as osteophyte formation, narrowed joint gap, and osteosclerosis; (3) history of fracture in the proximal femur, tumours or any other concomitant lower extremity diseases; (4) previous history of any surgical treatment in terms of core decompression, bone grafting, titanium implantation, or osteotomy; (5) previous history of any conservative treatment, such as extracorporeal shock wave therapy, hyperbaric oxygen, and alendronate; (6) inflammatory arthritis, including rheumatoid arthritis, suppurative arthritis and gouty arthritis; (7) having received steroid treatment in the last 6 months; and (8) pregnancy.
After meeting the inclusion criteria, the participants provided informed consent, and their baseline characteristics were assessed. The hips were randomly allocated to receive core decompression (CD) + an autologous bone graft (BG) or core decompression + an autologous bone graft + BBC (CD + BG + BBC) by a randomization schedule, which was generated by computer-based block randomization.
Surgical technique
All surgical procedures were performed under continuous epidural anaesthesia. For core decompression of the femoral head, we first determined the optimal entrance point for drilling, and then a 1.5-cm incision was made at the level of the greater trochanter. A 3.0-mm diameter Kirschner wire (K-wire) was introduced into the necrotic area with the tip placed in the subchondral bone area approximately 2 to 3 mm from the articular cartilage. Next, a 10-mm diameter trephine was drilled through the K-wire to the necrotic region. A cylinder of bone from the femoral neck and head was obtained. The necrotic proximal part was eliminated, and the healthy part was used for BBC grafting. The necrotic tissue remaining in the femoral head was removed with the bone curette. All the above steps were performed under C-arm X-ray guidance.
For bone marrow collection, a 50-mL syringe, heparinized in advance, was used to harvest the bone marrow from the posterior superior iliac spine. The bone marrow was centrifuged at 1500 revolutions per minute for 10 minutes in a bench-top centrifuge (Ependoff, AG 22331, Hamburg, Germany) with a sterilized chamber. The bone marrow was separated into three phases after centrifugation. We collected a total of 1 mL of bone marrow concentrate from the interface containing enriched bone marrow cells with a sterilized transfer pipette, and then the bone marrow concentrate was seeded on the cylindrical bone drop by drop to allow the cells to anchor onto the bone surface. A total of 10 μL of bone marrow concentrate was kept for cell counting after the surgery. The average number of bone marrow cells loaded onto the cylindrical bone was approximately 3 × 109 nucleated cells.
The bone graft with or without BBC was inserted and impacted into the necrotic region through the CD track with the guidance of C-arm X-ray. After surgery, the patients were instructed to remain non-weight-bearing for 4 weeks. Surgical complications were monitored after the operation.
Outcome assessment
Preoperatively, blinded evaluators collected baseline demographic information, including age, sex, aetiological factors, presurgical Ficat stage of ANFH, and location of the defect. Plain X-ray radiography and MRI were used to determine the Ficat stage of ANFH and the location of the necrotic lesions. All participants received follow-up at 24, 60, and 120 months postoperatively. Anteroposterior and frog-leg lateral radiographs were taken during each clinical assessment. The radiographic progression of ANFH was determined based on the Ficat classification system. The primary outcomes included visual analogue scale (VAS), Lequesne algofunctional index and Western Ontario and McMaster Universities Arthritis Index (WOMAC) osteoarthritis scores. The secondary outcome was the clinical failure rate of the operated hips at 5 years and at the time of the final follow-up. The clinical failure rate was defined as the proportion of hips that progressed to Ficat stage IV or that were subjected to THA.
Statistical analysis
The SPSS 22.0 software package (IBM Inc., USA) was used for statistical analysis. The data were described as median, maximum and minimum values. The chi-square test and Fisher’s exact test were used for categorical variables. The statistical significance of the demographic data and functional assessments between the two groups were determined by the Wilcoxon-Mann-Whitney test. Clinical survival was compared between each group with Kaplan-Meier survivorship analysis, and statistical significance was determined by the log-rank test. A Cox proportional hazards regression model was employed to detect the risk factors affecting the survival of the femoral head. The level of statistical significance for all tests was defined at P < 0.05.