At present, the treatment of ONFH in young and middle-aged patients is still controversial. Early diagnosis is of great importance, and early surgical intervention is effective for delaying the progress of necrosis and osteoarthritis[31-32]. Many surgeons are willing to perform hip-preserving surgeries for young and middle-aged patients with ONFH. However, no consensus has been reached on the effectiveness of so many hip-preserving surgeries[9,33]. The fibular grafting method used in this study is also one of the hot topics in this field of study .
During the period of necrosis and repair (ARCO stages II and III), stress could be concentrated between the necrotic and newly formed bones under the load-bearing condition, which would lead to a mild fracture of the bone trabecula, and affect the mechanical properties of the bone structure and repair of the necrotic area. Finally, the necrotic bone in the load-bearing area would collapse when the subchondral bone breaks. In view of the treatment of patients in these 2 stages, the following 4 problems must be solved to repair the necrotic area: (1) improvement of blood flow in the femoral head and promotion of regeneration of blood vessels; (2) effective removal of the necrotic bone; (3) reconstruction of the cartilage in the collapsed area of the femoral head to restore its shape and improve the matching relationship between the femoral head and the acetabulum; and (4) finally, improvement of the mechanical properties of the femoral head and prevention of its collapse.
CD combined with autologous or allogeneic fibular grafting can meet the 4 conditions well. The columnar supporting material, first proposed by Phemister in 1949 for the treatment of ONFH, was found to increase the rate of transformation of the structure into living bone and decrease the incidence of collapse. Some surgeons preferred autologous fibular grafting, while others were willing to perform allogeneic fibular grafting[20,25-26,35].
Recently, supporting materials mainly included 3 types, namely vascularized fibula, nonvascularized fibula, and tantalum rod. Vascularized fibular grafting requires a free peroneal artery and anastomosis with the lateral femoral circumflex artery[7,23]. This surgery can increase the blood supply in the femoral head, but the accompanying trauma is severe. Furthermore, the technical requirements are also high, and vascular embolism may even occur among a few patients in a short period after operation, which may conversely have a negative effect on creeping substitution. The tantalum rod is a porous tantalum metal with an elastic modulus similar to the fibula. However, in recent years, it was found that the host bone was unable to grow in the implanted titanium rod, resulting in further collapse of the femoral head[17,37]. In addition, the subsequent THA was more difficult. Ma et al reported that only 55 hips (52.9%) survived after porous tantalum rod implantation in the mean follow-up of 42 months. After retrieving the tantalum rod and femoral head for pathological and electron microscopic observation, Tanzer et al found that the failure of the tantalum rod implantation was related to minimal bone ingrowth. Therefore, the use of tantalum rods was also gradually reduced, and fibular grafting become the mainstream treatment[20,38]. Nevertheless, the choice between autogenous and allogeneic fibular grafting has been controversial in academia.
In this study, we performed CD combination with autologous or allogeneic fibular grafting. On one hand, CD can reduce the intraosseous pressure, stimulate the revascularization of the femoral head, and reconstruct the intraosseous circulation[39-40]. On the other hand, non-vascularized autologous and allogeneic fibular grafting can improve the mechanical properties[21-22,27]. The elimination of necrotic bone should be effective to decrease its distribution and accelerate new bone growth.
Not only were the postoperative HHS greatly improved when compared with the preoperative levels between the 2 surgical methods (Group A: from 65±7.2 to 80.3±14.5, p ˂ 0.05, Group B: from 66±5.9 to 82.4±13.6, p ˂ 0.05), but also the survival rates of patients without conversion surgery to THA were satisfactory at a mean postoperative follow-up of 13.1 years in group A and 10 years in group B (Table 2 and Table 3). In addition, the Kaplan-Meier analysis also revealed the high survival rate in groups A and B, indicating their great effect on the delaying the conversion to THA. Zeng et al retrospectively reviewed 18 patients with non-traumatic bilateral ONFH who underwent non-vascularized allogeneic fibular graft in one hip and, concurrently, one-stage THA on the contralateral side. They found that the overall survival rate of non-vascularized fibular allografting was 77.8% at a mean follow-up period of 53.3 months. Surprisingly, although the overall suvival rates were similar, the average time of conversion to THA in the group B was 4.1 years compared to 11.2 years in the Group A, which may result from the immunologic rejection and influence the repairing environment at early stage in the group B. The patients needed revision to THA were because of the collapse of femoral head and subsequent development of osteoarthritis. As for the autologous fibular graft, most studies focused on vascularized grafts for their reliable curative effect. Kawate et al reported that the overall survival rates (71 hips) could reach 83% at a mean follow-up of 7 years after free vascularized fibular grafting for the treatment of ONFH. They recommended that the degree of osteonecrosis should be less than 300° of the femoral head. Some studies reported that the survival rate with vascularized fibular grafting were higher than that with non-vascularized autologous fibular grafting, while others reported no significant difference between the two methods. Plakseychuk et al reported 86% survivorship in stage I and II ONFH after treatment with vascularized fibular grafting and only 30% survivorship after nonvascularized fibular grafting at the mean time of 7 years. However, in a study on large osteonecrotic lesions of the femoral head by Kim et al, no significant difference was found in the 3-year survival rate between vascularized and non-vascularized fibular grafts (p ˃ 0.05). Tetik et al demonstrated that although the clinical outcomes of vascularized fibular grafting were better than that of non-vascularized fibular grafting during the 1-year follow-up, no significant radiological difference were found between two procedures. Surprisingly, in spite of ARCO stage II or III, survival rate of 5-year follow-up was 100% and 10-year follow-up was 91.7% in group A. This rate of success in the early and mid-term has rarely been reported in other studies, ranging from 30% to 92.6%[21-22,26,32,38]. Keizer et al performed the non-vascularized fibular allografts for 60 hips, showing a clinical survival rate of 49% at 6 years and 38% at 10 years. In our opinion, the explanation of the results of high survival rates might be the following 3 surgical techniques: (1) Enough cancellous bone was grafted layer by layer and tightly impacted in all directions. (2) We supposed that the main effect of fibular graft on was its mechanical support but not the vascular implant, which was consistent with the results. (3) After surgery, all patients were demanded for non-weight bearing within the first 6 weeks and total weight bearing until the sixth postoperative month so that there was enough time for bone necrotic area to repair.
Compared with the allogeneic fibular graft, the autogenous fibular graft can better avoid immunologic rejection and save more medical resources and reduce surgical costs, but its use are constrained by the limited materials, longer operation time, and potential donor site complications. During operation, more attention should be paid to preserving the common fibular nerve. Although the allogeneic fibula was derived from the allogeneic bone and may cause immunologic rejection, its immunogenicity could be greatly reduced by bacterial inactivation, marrow removal, quick freezing, etc. Postoperative immune rejection symptoms can disappear in a short term in conjunction with intravenous administration of a low-dose glucocorticoid. In group B, 2 patients needed additional acetabular rim and femoral osteochondroplasty. This is a surgery for treating the femoroacetabular impingement when the ONFH progressed. During the surgery, we removed the abnormal bony structure of the femoral head and acetabular rim to improve the range of motion, which can delay THA again. There were no relative reports comparing this procedure in the two groups. According to the multivariate Cox model, we found no differences between HR of conversion to THA and etiologies. Yoo et al also found survival rates were associated with the patient's age but not etiology and stages of ONFH.
In summary, whether performing the autologous or allogeneic fibular grafting for young and middle-aged patients with ONFH of ARCO stage II or III is still a challenge for orthopedic surgeons because these patients present a highly heterogeneous group with extensively different symptoms, complaints, and expectations. Therefore, to achieve high postoperative satisfaction, a differentiated approach should be considered according to survival rates, clinical outcomes, bearable range of economic capability for patients, material cost, length of operation, postoperative immune rejection symptoms, and donor site complications (Figure 3).