Patients
A total of 61 consecutive ONFH cases (69 hips) of patients who underwent bone grafting through a window in the femoral head between April 2009 and March 2012 were prospectively investigated. The diagnostic criteria for ONFH, as identified with magnetic resonance imaging (MRI) were a low signal band in the T1-weighted image and a high signal band in the corresponding STIR sequence [35]. The inclusion criteria were: (1) consent to participate in the present study with at least 6 years of follow-up; (2) age ≤ 50 years; (3) no joint-space narrowing; (4) discomfort (pain in the hip, groin, buttock, or knee) that interfered with daily activity; (5) lack of a neurological disorder that could affect the source of the patient’s complaint; (6) no active connective tissue disease (e.g., rheumatoid arthritis, systemic lupus erythematosus); (7) ARCO stage II and III. Two patients (2 hips) were lost to follow-up and thus were excluded from the study. Ultimately, 59 patients (23 women and 36 men, 67 hips) were included in the analysis. The age of these patients ranged from 27 to 46 years old, with a mean age of 36.3 ± 5.3 years. The etiology of osteonecrosis was as follows: the use of corticosteroids in 20 patients, alcohol abuse in 37 patients, and idiopathic etiology in 10 patients.
All hips were graded according to guidelines provided by the Association Research Circulation Osseous (ARCO)[36]. Lesions that occupied < 15% of the femoral head were classified as Size A; lesions that occupied 15–30% of the femoral head were classified as Size B; lesions that occupied > 30% of the femoral head were classified as C [36]. The location of the lesion on coronal midsection T1-weighted images was classified as one of four types, according to criteria proposed by Sugano et al [37]. Type A accounted for one-third or less of lesions affecting the medial weight-bearing portion. Type B accounted for two-thirds or less. Type C1 accounted for more than two-thirds without extending laterally to the acetabular edge. Type C2 accounted for more than two thirds and extended laterally to the acetabular edge.
A total of 63 patients (72 hips), who underwent auto-iliac bone-grafting through a window at the femoral head-neck junction known as the “light bulb” approach for the treatment of osteonecrosis of the femoral head between March 2007 and April 2009, were retrospectively matched to 59 patients (67 hips) who underwent bone grafting through a window in the femoral head. The matching was based on the stage, extent, location, etiology of the lesion, average age, gender, and preoperative Harris hip score (Table I).
All cases were followed at 3, 6, and 12 months and then on an annual basis. Harris hip scores (HHS) and ARCO stage were recorded at each follow-up examination. Anteroposterior and frog-position X-rays, CT as well as MRI scans were obtained.
HHS was used to evaluate clinical outcomes. Excellent, good, fair and poor results were defined as > 90, 80–89, 70–79, and < 70, respectively. Scores < 80 points or patients who underwent THA were classified as examples of clinical failure. Clinical success was defined as a score of ≥ 80 points.
Each patient was also radiographically evaluated in terms of the progression in terms of the ARCO stage. Hips with > 3 mm of collapse or progress to ARCO Ⅳ were defined as radiographical failure [26].
Approval was obtained from the institutional review board, and written informed consent was obtained from each patient.
Surgical technique
Zhao et al performed modified trapdoor procedures. The patient was put in a supine position, and the ilium was elevated (10–15°) by placing a sandbag under the buttock. The procedure was performed using an anterior minimally invasive approach, and epidural or general anesthesia. An anterior straight incision, about 6 cm in length, was made about 1 cm distal and posterior to the anterior superior iliac spine. The hip joint capsule was exposed through the interval between the rectus femoris and the tensor fascia lata. The anterior part of the superficial aponeurosis was used to prevent damage to the lateral femoral cutaneous nerve. The femoral head was revealed by dissecting the joint capsule. MRI and computed tomography (CT) were preoperatively performed to localize the necrotic lesion and to determine the width of the tricortical bone block harvested from the iliac crest. The surgeon then created a window-like incision and used a scalpel to prepare a bone groove, extending from the head-neck junction to the acetabular rim, along the shaft of the femoral neck in the necrotic region of the femoral head. Necrotic bone was resected using osteotomes and power burrs, and curettage was performed until a bleeding surface was observed (Figure 1). Ipsilateral autologous tricortical iliac bone was trimmed into the shape, which was consistent with the length of the window and the depth of the necrosis. Finally, the iliac bone block and small cancellous bone graft were tightly filled by hammering at the surgical site. This restored the prototype of the articular surface. The screws were used to fix bone graft through the iliac bone to the femoral head, and the screw heads were hidden in the bone. The resected segment of articular cartilage was no longer covered. The range of joint motion was measured, as well as the stability of the labrum. Any cam impingement was corrected. The joint capsule was then closed, and the procedure was completed with fascial, subcutaneous, and skin stitches. A drain was left in place for 24 h (Figure 2).
Weight-bearing was forbidden during the first 3 months, after which patients were allowed to begin partial weight-bearing. Full weight-bearing began at least 4 months postoperatively, depending on the extent of the clinical and radiological union.
Statistical analysis
Continuous data are reported as means ± standard deviation. A paired t-test was used to compare changes in HHS. We used Pearson’s chi-squared or Fisher’s exact test to evaluate the relationship between variables and results. All analyses were performed with SPSS statistical software package, ver. 19.0 (SPSS, Inc., Chicago, Illinois, USA). P < 0.05 was considered to be statistically significant.