The relationships between bone marrow and osteonecrosis of the proximal femur are complex.4) Changes in the bone marrow signal are observed on magnetic resonance imaging (MRI) scans of patients with osteonecrosis. There is an increase in the amount of fatty marrow in the intertrochanteric portion of osteonecrotic hips. Abnormalities of osteogenic stem cells are present in the bone marrow of some of these patients. Steroids have been shown to produce adipogenesis and stimulate fat-specific genes in cloned bone-marrow cells; following corticosteroid therapy in osteonecrotic patients, abnormalities have been demonstrated in the bone marrow of the iliac crest, with a decrease in the stem cell pool.5) The effects of steroids on cloned bone marrow cells in vitro include the production of adipogenesis and the stimulation of fat-specific genes.

As a consequence, intramedullary vascularity is altered and this may be a predisposing factor for osteonecrosis since changes in bone marrow and bone remodeling are linked. Another consequence is the lack of osteogenic cells, which could influence two different events in the pathogenesis of osteonecrosis; the occurrence of osteonecrosis itself and the bone repair occurring after osteonecrosis. A decrease in osteogenic stem cells in the femoral head has been observed beneath the sequestrum and in the intertrochanteric region. This has also been confirmed by the observation of the extent of osteocyte death in the proximal femur seen in patients who underwent total hip replacement (THR) for osteonecrosis. One of the reasons for insufficient creeping substitution in bone remodeling after osteonecrosis in the femoral head may be the small number of progenitor cells present in the femoral head in these patients. Although both research and clinical studies have shown that dead bone may be replaced by living bone, the osteogenic potential for repair is low in osteonecrosis. Bone marrow activity5) in the proximal femur of patients with corticosteroid-induced ONFH was evaluated and compared with a control group without ONFH. A decrease in the number of MSCs was found outside the area of ONFH in patients with a corticosteroid induction. Reconstruction and repair has been observed after core decompression, but is usually incomplete. Grafting with autologous bone marrow could lead to increased repair in osteonecrosis of the hip.

Autologous bone marrow transplantation was therefore proposed for the treatment of osteonecrosis. The effectiveness of bone marrow mononuclear cells may be related to the availability of stem cells endowed with osteogenic properties arising from an increase in the supply of such cells to the femoral head through bone marrow implantation.678910) Another possible explanation for the therapeutic effect of bone marrow implantation is that injected marrow stromal cells secrete angiogenic cytokines, resulting in increased angiogenesis and subsequent improvement in osteogenesis. Finally, bone marrow-derived mononuclear cells are able to elicit the formation of new blood vessels by the presence of endothelial cell progenitors or hemangioblasts in this cell fraction. This may be due both to the supply of progenitor cells and to angiogenic cytokines produced by bone marrow cells. Endothelial progenitors can actively engage in vasculogenesis in tissue devoid of vessels, and in neoangiogenesis from the pre-existing capillaries. Besides the generation of new capillaries, the growing endothelia enhance mobilization and growth of mesenchymal progenitors through the angiopoietin 1-Tie2 pathway, which generates pericytes and vascular mural cells required for new vessel growth and stabilization. A broad capacity of differentiation of perivascular mesenchymal cells has been shown, and participation of perivascular mesenchymal progenitors in repair of adjacent tissues has been described both in experimental models and humans. In addition, local ischemia that activates HIF1A signaling and mobilization of circulating progenitors through the SDF1-dependent pathway, may supply permanent stimuli for blood vessel repair and a supply of new cells for bone regeneration.

Adult MSCs represent usually a heterogeneous population of cells, with a positive immunophenotype for STRO-1, CD73, CD146, and CD106 and a negative one for CD11b, CD45, CD34, CD31, and CD117.678910) MSCs act via multifaceted pathways not yet completely understood in order to augment regeneration, and includes mechanisms that mediate homing of administered MSCs to sites of injury. However, two main inherent functions of MSCs can be distinguished. The first is the secretory or "trophic" function of MSCs, which includes the secretion of a wide spectrum of factors with immunomodulatory, anti-inflammatory, antiapoptotic, proangiogenic, proliferative, or chemoattractive capacities, among others. Second, administered MSCs are thought to orchestrate the differentiation process together with differentiated or undifferentiated resident cells for functional tissue restoration.

Given the different modes of MSC delivery as well as the time and efforts necessary for performing meaningful randomized controlled clinical trials, the search for an optimized procedure has remained elusive. However, several studies have been conducted that use mesenchymal stromal cell-based approaches in the clinical setting, and some of the basic principles of stromal cell delivery are now outlined. A successful regenerative approach based on the use of autologous material usually comprises the harvest of cells either derived from bone marrow by, for example, aspiration from the iliac crest or by surgical removal from a donor site.

Hernigou and colleagues1112) pioneered the clinical application of a cell-based strategy for the treatment of AVN, percutaneously injecting an autologous bone marrow concentrate into the necrotic area of femoral heads. The strategy is driven by the hypothesis that injected bone marrow cells can repopulate the trabecular bone structure and subsequently revitalize and remodel the necrotic bone. For this technique, a bone marrow aspirate (BMA) is harvested from the iliac crest and the mononuclear cell fraction is isolated. After core decompression, the cell suspension is injected into the necrotic area.

Hernigou and Beaujean3) have published clinical data on their experience. They treated 189 hips in 116 patients with autologous BMCs and had a follow-up of 5 to 10 years. Satisfactory results were achieved in the majority of patients according to improvement of the Harris hip score, radiographic assessment and refusal of THA. The prognosis was not only highly related to the stage of disease, but also to the progenitor cells transplanted. When patients were operated on before collapse and received a greater number of BMC injections, a better outcome could be expected. In 2008, Hernigou et al.16) retrospectively analyzed 534 hips in 342 patients with ONFH treated with autologous BMC transplantation. The results were really encouraging. They showed that the volume of necrosis would decrease from 26 cm³ to 12 cm³ in 371 patients with an average follow-up of 12 years. There were only 94 patients who progressed to THA. The author concluded that the best indication for cytotherapy of ONFH was in the precollapse stage when the hip was symptomatic; and in some patients with Steinberg stage III, ONFH successful outcome could be obtained in 5 to 10 years.

In another report from Gangji and colleagues,1718) patients with ONFH were treated by injection of bone marrow stromal cells (BMSCs). Osteoprogenitors and osteoblasts from bone marrow were separated and expanded in vitro, and injected into the necrotic zone after differentiation under autologous conditions. Pain reduction, necrotic lesion decrease and functional improvement were recorded in the early period, and only minor side effects were found.

Papakostidis et al.19) and Hernigou et al.20) investigated through literature review whether implantation of autologous BMA containing high concentrations of pluripotent mesenchymal stem cells into the core decompression track would improve the clinical and radiological results as compared with the classical method of core decompression alone. The primary outcomes of interest were structural failure (collapse) of the femoral head and conversion to THR. They identified 496 citations between 2002 and 2015, and selected papers of interest.21222324252627282930) Their findings indicate that the application of autologous bone marrow concentrate (autologous cell therapy) in combination with core decompression in osteonecrotic femoral heads is superior to core decompression treatment, as it was found to markedly decelerate the progression of the disease to the stage of femoral head collapse, and also limit the need for THA. Some component studies also reported on clinical results, demonstrating that autologous cell therapy in addition to core decompression for the treatment of ONFH resulted in reduction of painful joint symptoms and improvement in the Harris hip score as compared with sole use of the core decompression technique.

Clinical outcomes are usually analyzed with hip score, collapse, and the need for THA. The reduction of volume3132) and restoration of original MRI signals of a living bone marrow is sometimes observed. However, it appears that there are many variations of the signal when abnormal tissue (normal tissue, ossification, fibrous tissue) is repaired with absence of collapse; furthermore, when scaffolds are used (hydroxyapatite), the presence of bone substitute remains visible in the femoral head for a long time (several years), and of course acts as an artifact to evaluate the exact repair. Therefore, at this moment, most of the clinical studies keep as an outcome the absence of collapse during the evolution.

The number of MSCs in a normal femoral head was evaluated by Hernigou et al.5) and Homma et al.33) by bone marrow aspiration and femoral head fragmentation. Bone marrow was collected by aspiration from the femoral head of patients receiving THA. The data showed that the total number of MSCs present in 1 cm³ of a femoral head was on average of 700 ± 264 MSCs per cm³. Since the femoral head has an average volume of 50 cm³, a total of 35,000 MSCs may be considered as a useful approximation of the number of MSCs present in a femoral head. This number may be considered as the target number to load in an osteonecrotic femoral head to reestablish the same number of MSCs as in a normal femoral head. The influence of the number of progenitors on the results of treatment of hip osteonecroses of the femoral head was evaluated in patients treated through grafting with autologous bone marrow. MRI observation in a series of osteonecroses demonstrated that the volume repair was on average 15 cm³ with an injection of a bone marrow graft of 20 mL containing approximately 2,500 progenitors per mL. In adult bone remodeling, these processes of bone formation take place in the context of the basic multicellular unit described by Parfitt et al.34) and Frost.35) Osteoblasts begin secreting matrix within a day, and matrix synthesis increases over several days to a maximum rate of approximately 1.5 µm per day over an area of approximately 150 µm² per osteoblast, resulting in synthesis of approximately 225 µm³ per day per osteoblast. It is estimated that the volume of bone matrix made by one osteoblast is approximately 5,000 µm³. The osteocyte density is reported to be 0.000047 osteocytes/µm³ in cancellous bone. As a first approximation, based on a mean bone matrix of 33% in cancellous bone, one can estimate the number of osteocytes in 1 cm³ of cancellous bone to be in the range of 20 million. There are approximately 20 million osteoblasts or osteocytes per mL of new bone. Since there are approximately 2,500 progenitors per mL of prepared marrow graft, each must have divided a minimum of 12 or 14 times to obtain 1 mL of new bone, assuming that all the bone marrow graft retained the ability to make trabecular bone (2,500 × 2¹⁴ = 20 million osteoblasts). Histologic observations have demonstrated that the ratio of trabecular bone is 1/3 in the femoral head, the others 2/3 being fat and hematologic cells. Assuming that in the repair of 15 cm³ in the femoral head that 1/3 is bone, this means that 5 cm³ of bone were obtained.

In future research, several questions will need to be addressed. For example, is the differentiation potential of MSCs from different sources (bone marrow, fat, and periosteum) the same? Are their functional abilities after repeated culture the same? How different are bones formed by implanted MSCs from normal bones in terms of histology and biomechanics? In addition, the risk of forming cancer at the implanted site should be evaluated. Hernigou and colleagues4647) found that patients treated with cell therapy do not have a greater incidence of cancer than the rest of the population. They analyzed the occurrence of cancers by follow-up, cell number, sites of cancer, age, gender, and the pathology that was treated, and found that the risk of cancer was not increased in patients with longer follow-ups or in patients who had received a higher number of MSCs. However, this study was performed with autologous MSCs from bone marrow; there are no reports of implantation with allogenic MSCs or expanded MSCs with long-term follow-up.