Main findings
This was the first report and long-term outcome of autologous PBSCs transplantation combination with mechanical support for treating ONFH. A latest clinical study reported intra-arterial infusion of auto-PBSCs for treating ONFH,the follow-up time was 53.96 ± 21.09 months (ranged from 8 to 84 months), which was the longest one [23]. The implantation of PBSCs and operation were different from this study. The main findings of our study as follow,1. Autologous PBSCs transplantation improved the results of treatment ONFH with core decompression and allogeneic fibula grafting; 2. The autologous PBSCs were rich in peripheral blood and could be was mobilized by rhG-CSF and astragalus injection which reduced the time and dose of exposure to rhG-CSF; 3. It was easy to reach a high enough concentration of PBSCs for collection with less trauma and lower cost; 4.The combination option for treating ONFH was safe and efficacious.
Core decompression and allogeneic fibula grafting
The option for treatment of ONFH was categorized into nonoperative and operation. There were many nonoperative treatments available for ONFH, most with limited supporting data[24]. Operation including hip-preventing surgery and THA was appropriate for most ONFH except for asymptomatic hip with small, medially located lesions[25]. THA was indicated for advanced-stage osteonecrosis in older patients or those who have failed joint-preserving treatment with excellent results in terms of pain relief and survivorship [24]. However ONFH often occurs in young patients who will receive revision THA in future if received THA too earlier. For young patients with early stage ONFH hip-preventing surgery was eligible. Core decompression was the first hip-preventing surgery and has beneficial results, bur the mechanical properties of the bone was inferior after core decompression [25]. It aimed to improve blood flow to the femoral head by decreasing intraosseous pressure [23] and the prognosis in bone healing[26].It was indicated for early-stage ONFH resulted from only venous stasis without artery blood supply insufficiency [2].Postoperative fracture and further collapse of the articular occasionally occurred because of increasing strains of the bone in the vicinity of the holes, which was more pronounced in the bone with necrotic material properties in the femoral head, whatever the number, location and diameter of the drilling holes[3, 4] .Early study reported that the collapse rate was 78% at 24 months postoperatively, it had no greater value than conservative management in preventing collapse in early osteonecrosis of the femoral head[27]. In order to prevent femoral head collapse by providing structural support, different material such as autogenous bone, allograft bone and bone graft substitute were utilized to filling the defect to increase the proximal femur strength[28, 29, 30, 31]. Core decompression and allogeneic fibular grafting has been used in our hospital for many decades, the mid-long term outcome was excellent[32, 33]。Modified non-vascularized allogeneic fibula grafting combined with core decompression and bone grafting also was an effective and cost-effective treatment on early femoral head necrosis with satisfactory survivorship and could improve the clinical outcomes, delay the disease progression, and enhance the quality of life for patients[34].The fibular allograft procedure had favorable biomechanical efficacy and biological repair ability, it can prevent the necrotic femoral head from collapse in a certain extent and improve the survival rate,and its clinical effect is obviously better than pure conservative treatment even as good as vascularized fibular grafting[35].
Bone marrow stem cells transplantation
Stem cells are a group of cells with the ability to self-renew and differentiation into target cells. In adult, mesenchymal stem cells (MSCs) were the main stem cells which consist in bone marrow called bone marrow stem cells (BMSCs) and adipose tissue called Adipose-derived MSCs (aMSCs). A study in vitro found that aMSCs outperform BMSCs in growth rate and bone differentiation potential in the setting of osteonecrosis, suggesting they may provide a more-potent regenerative therapeutic strategy in ONFH[36]. Autologous aMSCs transplantation improved osteogenesis and the microstructure of vascular deprivation-induced osteonecrotic tissue in rabbit[37]. However aMSCs was seldom utilized for treating ONFH clinically [38]. BMSCs had been used for treating ONFH in 1990 combined with core decompression[39]. The safety and biodistribution of BMSCs in bone had been confirmed by an animal research[40]. Recently several reviews reported that autologous BMSCs combination with core decompression had better pain relief and clinical outcomes compared with core decompression alone in the treatment of early-stagy ONFH, which can delay the collapse of the femoral head more effectively[17, 41, 42]. Another matched pair case-control study suggested that implantation of MSCs into the femoral head at an early stage of ONFH lowered the THA conversion rate, however ARCO stage progression is not affected by this treatment[43]. The aforementioned studies did not fill the defect followed core decompression to provide structural support, this may influence the outcome. Tao Wang,et al performed a thorough debridement of all necrotic lesion, pack autogenous cortical and cancellous bone which were harvested from the ipsilateral iliac crest tightly into the femoral head, implanted bone-marrow mononuclear cells containing mesenchymal stem cells into the necrotic lesion for patients with ARCO II-III ONFH, the overall clinical success rate is 80% without infection, femoral neck fracture or other complications[44]. Dewei Zhao described a modified technique using bone marrow mesenchymal stem cells (BMMSCs) associated with porous tantalum rod implantation combined with vascularized iliac grafting for the treatment of end-stage ONFH, the outcome showed that Harris hip score improved significantly from 38.74 ± 5.88 points (range 22–50) to 77.23 ± 14.75 points (range 33–95) and the joint-preserving success rate of the entire group was 89.47% for ARCO stage IIIc and 75% for ARCO stage IV, which suggested than this intervention was safe and effective in delaying or avoiding total hip replacement for end-stage ONFH[45].
However, a considerable number of patients who received BMSCs transplantation did not achieve a satisfactory outcome in terms of repair of the femoral head necrotic area. One reason was that the success of hip joint preserving depended on the stage and amount of osteonecrotic lesion[46].Patients with a large sized lesion or medium sized laterally located lesion achieved poor outcomes[43]. Another reason may be associated with the quality of implanted stem cells, which decided the ability of differentiation into bone[12, 47]. Whereas cellular activity and ability to differentiate into bone of BMSCs in patients with Corticosteroid-Induced ONFH have decreased, which may be a factor in why patients with corticosteroid-induced ONFH treated with autologous iliac crest bone marrow grafting progress to collapse [12]. Other cell sources need consider.
Peripheral Blood Stem Cell Mobilization
Under steady-state conditions less than 0.05% of the white blood cells (WBC) are PBSCs, mobilization of CD34 + cells from the bone marrow into the peripheral blood was necessary [48, 49] .G-CSF was the most commonly used with few side effects and high mobilization efficacy[50]. If PBSCs were poor mobilization in patients with hematological malignancies, plerixafor could be added to enable rapid and efficient mobilization[51] ,which was a salvage strategy in dose of 0.24 mg/kg to haploidentic healthy donors[50].There were no consensus about the optimal time and dose of G-CSF to mobilize PBSCs. In the early study period, leukapheresis for PBSC collection was performed for 2–3 days in 75 (37.9%) patients, 4–5 days in 60 (30.3%) patients, and 6–11 days in 63 (31.8%) patients[52]. At present 4–5 days mobilization was commonly used[19, 50, 53, 54].The dose of rhG-CSF was 10–16 µg/kg/day with low rates of short-term serious side effects and no potential long-term complications[50].
In our study patients received rhG-CSF (5ug/kg/d) subcutaneous injection and astragalus injection intravenous drop for 3–4 days. On the third day WBC count was test. when the WBC count > 30 × 10e9/L the mobilization stopped, if ༜30 × 10e9/L, the mobilization continue more 1 day. Autologous PBSCs were collected by COM. TEC( Fresenius,Germany) on the next day. 70% patients can avoid one more day of G-CSF exposure and its potential side effects. The dose we used was lower than current literature reports, astragalus injection intravenous drop maybe the reason. Astragalus polysaccharide was the main component of astragalus, which enhanced the secretion of granulocyte-macrophage colony stimulating factor (GM-CSF) in vitro[55, 56].However the synergistic effect of both was unclear, need further research.
Cell volume was important for success of PBSCs transplantation, higher dose of CD34 + PBSCs is associated with better survival. Generally, the minimum quantity of PBSCs for one transplant is defined as 2.0 × 106 CD34 + cells/kg bodyweight [57].Optimal dose of CD34 + cell for autologous and allogeneic was 5 × 10e6 and 4.5 × 10e6/kg respectively [58]. The forementioned cut-off was intravenous graft for multiple myeloma or leukemia. Local graft was commonly performed for treating ONFH. It was reported that stem cell therapy for treating ONFH, the injection of 106 to 109 cells may be reasonable, however, the optimal number still needs to be investigated[59]. Mao Q et al infused 1.71 ± 0.7 × 106 PBSCs into the femoral head through the medial circumflex artery[21]. In this study we infused 0.87 (0.64–1.1) X108/ml CD34 + PBSCs 10 ml into medullary cavity of the fibular and then flowed directly into femoral head.
So PBSCs in peripheral blood (PB) must be enough before leukapheresis. Some factors had been used to predict mobilization. CD34 + cell level at baseline was the most used. The basal CD34 + either ≤ 2 cells/µL or ≥ 3 cells/µL well predicted the levels of CD34 + on day 4, which determine whether mobilization should be attempted or not[53]. Whereas in adult it was set that PB-CD34+ count > or = 15 CD34(+) cells/microL and ≥ 20 cells/µL were the best predictor to begin the apheresis procedure [60, 61]. However sometimes the number of CD34 + cells in PB may be persistently lower than the threshold at which PBSCs collection begins in poor mobilizers, and therefore, it was difficult to decide the timing to start PBSC collection[62]. In addition CD34 + cells test was expensive. Many institutions still consider WBC count in PB as one of valid surrogates. However, the optimal WBC count in PB to start PBSC collection has not been systematically tested. Sung KW et al proposed that WBC count in PB exceeded 4,000/µL predicted a higher CD34 + cell yield in Children with high-risk solid tumors[62]. Lydon H et al found that the mean WB count of 16.1 ± 5.0 × 106 /ml was sufficient for apheresis in ovine model[63]. In our study we set the cut-off of WBC as ༞or = 30 × 109/L, and collected sufficient PBSCs.
PBSCs transplantation
PBSCs are mononuclear cells enrich in peripheral blood, which was not fully accepted until 2004, when Valenti MT,et al found that human PBSCs circulated among mobilized peripheral blood[64].PBSCs had similar characteristics and chondrogenic differentiation potential to BMSCs[65], also can differentiate into osteoblasts and adipocytes [19, 66], neuron-like cells[67]. In clinical practice, intracoronary infusion PBSCs improved cardiac function and promoted angiogenesis in patients with myocardial infarction[68]. Animal experiment and clinical study had demonstrated that PBSCs also could be used for treating ONFH[20, 69, 70]. The possible mechanism were that PBSCs improved blood supply of the femoral head[69]and enhanced the bone regeneration by up-regulating BMP-2 expression and down-regulating PPAR-γ expression, both of which were the most typical and closely relative genes in bone regeneration[70] .
To our knowledge this was the first report of clinical outcome study of a combinational treatment for ONFH with autologous PBSC transplantation combination with core decompression and allogeneic fibula grafting. Mao Q reported a randomised controlled clinical trial in which tantalum rod implantation in combination with targeted intra-arterial infusion of PBSCs was utilized for treatment of ONFH[21]. There was some difference from our study. Firstly the porous tantalum rod was utilized for mechanical support, secondly PBSCs were infused through medial circumflex femoral artery into femoral head. There was no comparative study on porous tantalum rod and allogeneic fibula grafting. In Mao Q’ s study PBSCs were intra-arterial infused, the fate of PBSCs after transplantation was not examined, the cell homing was unclear. In our study PBSCs were infused into medullary cavity of the fibular and then flowed directly into femoral head, which could existed in the bone tunnel after core decompression for 6months[71] when PBSCs differentiated into osteoblasts[20] and endothelial progenitors with angiogenic properties containing in the PB mononuclear cells (MNCs) promoted MSCs produce proangiogenic factors as well as differentiate into mural pericytes[72]
Limitations
There were some limitations of this study. Firstly this was a retrospective study, not a prospective randomised controlled clinical trial. Secondly the case number was small, we only included patients between February 2012 to February 2013.Thirdly how PBSCs playing a part in the repair of ONFH was not investigated, but other research explained the possible mechanism. Forth the synergistic effect of rhG-CSF and astragalus injection was unclear, need further research.