Cytotherapy is an insufficient method for promoting bone repair in steroid-associated osteonecrosis (SAON), and this has been attributed to impairment of the bioactivity of bone marrow–derived stem cells (BMSCs) after pulsed administration of steroids. Cryopreserved autologous bone marrow–derived mononuclear cells (BMMNCs), which contain BMSCs, might maintain their bioactivity in vitro. This study sought to investigate the effects of cryopreserved BMMNCs, before steroid administration, on the enhancement of bone repair in an established rabbit model of SAON.
For in vitro study, bone marrow was harvested 4 weeks before SAON induction from the iliac crests of rabbits (n = 10) to isolate fresh BMMNCs, and the BMMNCs were then cryopreserved for 8 weeks. Both the fresh and the cryopreserved BMMNCs were evaluated for their bioactivity and osteogenic differentiation capacity. In addition, BMMNCs were isolated 2 weeks after SAON induction and subjected to the same evaluations. For in vivo study, cryopreserved BMMNCs were implanted into the bone tunnel during core decompression of the femur (n = 12 rabbits) after the induction of SAON, and tissue regeneration was evaluated by micro–computed tomography and histologic analyses at 12 weeks postoperation.
In vitro, there were no significant differences in the bioactivity or ability to undergo osteogenic differentiation between fresh BMMNCs and cryopreserved BMMNCs, but after SAON induction, both features were decreased significantly. In vivo, the bone mineral density, ratio of bone volume to total volume of bone, and volume and diameter of neovascularization within the bone tunnel were significantly higher in the BMMNC-treated group compared to the nontreated control group at 12 weeks postoperation.
Cryopreserved BMMNCs maintained their bioactivity and promoted bone regeneration and neovascularization within the bone tunnel after core decompression in this rabbit model of SAON.
Steroid-associated osteonecrosis (SAON) is one of the serious complications that occurs after pulsed administration of steroids in several medical conditions, including rheumatic diseases, such as systemic lupus erythematosus, as well as organ transplantation and severe acute respiratory syndrome (1–4). Recent advances in the understanding of the pathophysiology of SAON have shown that bone repair at the early stages of this disorder is inadequate, due to the decreased activity of the bone marrow–derived stem cell (BMSC) pool (5–9). Core decompression is one of the least invasive surgical procedures indicated for the early stages of osteonecrosis (ON), when the ON lesion is still small (10–13). However, the postsurgery prognosis is rather poor, and this is attributed to incomplete reconstructive repair and weakening of the trabecular bone within and next to the necrotic region that may, ultimately, progress to subchondral collapse (11, 14). One of the reasons for incomplete reconstructive repair after SAON is that the number of progenitor cells is decreased, especially in the proximal femur (7, 8, 15).
Bone marrow contains abundant stem cells that may differentiate into regenerative precursor cells, such as osteogenic and angiogenic precursors. BMSCs, considered an appropriate source of precursor cells, have been implanted for the treatment of ON lesions after core decompression at the early stages of ON (16). However, the limitation of such an approach is that implantation of autologous bone marrow–derived mononuclear cells (BMMNCs), which contain BMSCs, has been less promising as a treatment for SAON than for other types of ON, according to the results of recent trial experiments (15, 17, 18). This may be largely attributed to the decreased activity of the implanted marrow progenitor cells under the systemic toxic influence of previous pulsed administration of steroids (5, 6, 9). It was reported that cryopreserved autologous BMMNCs had a similar osteogenic and angiogenic differentiation capacity as that of fresh autologous BMMNCs, and could be implanted without the need for additional treatment in vitro (17, 19). Furthermore, such a capacity remained during the various cell-growth passages in culture. Therefore, a cell-based strategy involving the implantation of cryopreserved autologous BMMNCs prior to pulsed steroid administration might facilitate the reparative processes of osteogenesis and angiogenesis that are needed for reconstructive repair in SAON.
In order to investigate the above-described strategy for potential translation into clinical application, we used a rabbit model of SAON that has been previously established by our group. In our previous studies, we demonstrated that this may be a useful model for assessing bone repair when pulsed administration of steroids has delayed the bone healing after core decompression (9, 20). Accordingly, the present study was undertaken to evaluate the use of cryopreserved autologous BMMNCs to promote the repair of ON lesions, using this established rabbit model of SAON.
MATERIALS AND METHODS
In vitro study.
Ten 28-week-old male NZW rabbits were used for establishing the SAON model. ON was induced using a previously published inductive protocol, involving a single injection of 10 μg/kg of lipopolysaccharide, followed by 3 injections of 20 mg/kg of methylprednisolone (21, 22). This protocol has been shown to induce a high incidence of ON of up to 93%, and to result in low or no mortality. The formation of ON lesions can occur as early as 2 weeks postinduction (21, 23, 24). The experimental protocol was approved by the Animal Experiment Ethics Committee of the Chinese University of Hong Kong.
Four weeks before and 2 weeks after the induction of ON, rabbits (n = 10) were placed under general anesthesia with xylazine (20 mg/kg body weight) and ketamine (50 mg/kg body weight), and the bone marrow was harvested from both iliac crests. Briefly, under strictly sterile conditions, a beveled metal trocar of 6 cm in length with a bore of 1.5 mm was pushed, by hand, deep into the iliac crest. The bone marrow was aspirated with 10-ml syringes that were rinsed with 1,000 units of heparin. The 5 ml of bone marrow obtained from each iliac crest was collected and mixed with an equal volume of balanced salt solution (final volume 20 ml).
BMMNCs were isolated from the marrow aspirate by gradient centrifugation at 435g for 30 minutes at room temperature, using Ficoll-Paque Plus (1.077 gm/ml; GE Healthcare Biosciences) as reported previously (25). The density of the mononuclear cells was determined in a hemocytometer using the trypan blue exclusion method, and more than 5 × 107 BMMNCs were obtained from each animal. Fresh BMMNCs and cryopreserved BMMNCs without steroid treatment, as well as steroid-treated fresh BMMNCs, were used for comparing their bioactivity in vitro.
Cryopreservation and thawing of BMMNCs.
After isolation of BMMNCs from the bone marrow 4 weeks before the induction of ON, the mononuclear cells were resuspended in freezing reagent, which included 10% DMSO plus 90% fetal bovine serum (FBS; Gibco) (26), and then cryopreserved in freezing tubes containing liquid nitrogen, at a density of 107 cells per tube. After 8 weeks, the cells were quickly thawed in a water bath at 37°C, followed by suspension in culture medium consisting of Dulbecco's modified Eagle's medium (DMEM; Gibco), 10% FBS, 100 units/ml penicillin, and 100 gm/ml streptomycin.
Assessment of cell viability and colony formation after freezing.
The BMMNCs were thawed and the total recovery of the cells was assessed by counting the cells with a hemocytometer; the viability of the cells was determined using the trypan blue exclusion method. The colony-forming unit (CFU) assay was used for enumeration of adherent colonies (25). The BMMNCs were seeded in two 6-well culture plates at a density of 1× 105 per well, and then cultured in a humidified incubator at 37°C and 5% CO2. The culture medium was changed on day 3 after initial plating to remove the nonadherent cells, and thereafter the medium was changed every 3 days. Ten days after seeding of the BMMNCs, the adherent colonies, i.e., those in which the number of adherent cells was >30 as determined by staining with 0.5% Giemsa staining solution for 30 minutes, were counted in one of the 6-well plates (8). The results are expressed as the mean number of CFU per 106 BMMNCs, as well as the mean diameter of the CFU. In another 6-well plate, cultures were digested using 0.25% trypsin and the cells were counted using a hemocytometer. The results are expressed as the mean number of adherent cells per 106 BMMNCs. For comparisons, fresh BMMNCs before steroid administration were compared with fresh BMMNCs after steroid administration.
Assessment of osteogenic differentiation of BMMNCs after freezing.
Cells from the same freezing tube were seeded onto 24-well plates at a density of 1 × 106 per well and cultured with 10% FBS–DMEM. After determining the confluence of colony-forming cells in each well, the osteogenic differentiation of these primary cells was induced with the use of DMEM supplemented with 0.1 μM dexamethasone, 50 μM ascorbate-2-phosphate, and 10 mM β-glycerophosphate (27). The osteogenic medium was changed twice a week.
Osteogenic differentiation was confirmed by the expression of alkaline phosphatase (AP), detected using staining for AP and the measurement of AP activity at 5 days after induction of ON. AP activity, expressed as nmoles p-nitrophenol/minute/μg protein, was determined with p-nitrophenylphosphate as the substrate, with measurement of kinetics performed for 10 minutes (in 1-minute intervals) at 405 nm using a universal microplate reader. For statistical comparisons, AP activity was normalized to the protein content, with results expressed as nmoles p-nitrophenol/minute/μg protein. The subsequent osteogenic differentiation of the cells was detected by staining for calcium nodules, using 2% alizarin red S. The results of staining demonstrated the deposition of calcium matrix at 4 weeks after induction. The extracellular calcium deposit was evaluated using a previously reported method (27, 28). The Ca2+ ions in the extracts were measured using a calcium colorimetric assay kit (K380-250; BioVision), with results expressed as μg calcium/μg protein. The osteogenic differentiation ability of the fresh BMMNCs was compared before and after steroid administration.
In vivo study.
Twelve 28-week-old male NZW rabbits were used for the in vivo experiments. BMMNCs were collected at 4 weeks before SAON induction and cryopreserved in a freezing tube with liquid nitrogen. Rabbits were placed under general anesthesia with xylazine (20 mg/kg body weight, intraperitoneally [IP]) and ketamine (50 mg/kg body weight, IP) at 2 weeks after SAON induction. Thereafter, a 2.0-cm longitudinal incision was made at the medial aspect of the distal femur, and a bone tunnel at the epiphysis was made using a drill with a diameter of 3.0 mm, from the attachment of the medial collateral ligament to the contralateral cortex, parallel to the coronal plane of the knee joint. Using this procedure, a bone tunnel was created at the coronal plane of the distal femora bilaterally (Figure 1, panels A1 and A2). We have previously confirmed that formation of ON lesions and impaired healing are characteristics found in the distal femur after core decompression in this rabbit model of SAON (20).
The cryopreserved BMMNCs were thawed, and ∼5 × 107 cells were directly implanted into the bone tunnel after core decompression. The tunnel was blocked with bone wax before closing the skin wound via suturing. Bone tunnel without implantation of BMMNCs served as the nontreated control. Analgesic (Temgesic, 0.02–0.05 mg/kg body weight, subcutaneously) was given every 6–12 hours for up to 3 days after surgery. At 12 weeks postoperation, micro–computed tomography (micro-CT) and histologic analyses were performed for evaluation of osteogenesis and repair within the bone tunnel.
Micro-CT analysis of new bone formation in the bone tunnel.
Twelve weeks after core decompression and implantation of the BMMNCs, the animals were anesthetized with xylazine (20 mg/kg body weight, IP) and ketamine (50 mg/kg body weight, IP). The left and right femora of all rabbits (n = 12) were prepared for micro-CT scanning of the distal femur (XtremeCT; Scanco Medical), with a spatial resolution of 40 μm. The bone mineral density (BMD) and trabecular architectural parameters, including the bone tissue volume density (bone volume/total volume of bone [BV/TV], expressed as a percent), connectivity density (Conn.D; expressed as 1/mm3), trabecular number (Tb.N; expressed as 1/mm), trabecular thickness (Tb.Th; expressed as mm), and trabecular separation (Tb.Sp; expressed as mm), were quantified for the newly formed bone within the bone tunnel.
After micro-CT scanning, half of the rabbits (n = 6) were killed using 30% pentobarbital (1 ml/4 kg body weight, IP), and the femoral samples were harvested for histologic analysis. On day 14 and day 4 before the mice were killed, the fluorescence tracers calcein green (5 mg/kg; Sigma) (at 14 days) and xylenol orange (90 mg/kg; Sigma) (at 4 days) were injected intramuscularly for labeling the newly mineralized bone (29). The samples were cut into 2 parts sagittally to the bone tunnel, and the lateral parts were decalcified using 9% buffered formalin acid, before being embedded in paraffin and sectioned longitudinally at a thickness of 7 μm, followed by hematoxylin and eosin staining. The newly formed bone and collagen alignment within the bone tunnel were observed using a light microscope (Q500MC; Leica) with polarized light.
The medial parts of the samples were immersed in 10% formalin for 48 hours and then embedded in methyl methacrylate resin (30). The cross-sections of resin blocks were prepared to a thickness of ∼50 μm and then observed under a fluorescence microscope (DM IRB; Leica) with a Moticam 2300 camera. The single-labeled (sL) surface, double-labeled (dL) surface, and interlabeled thickness (IrLTh) were measured, and the data were used for the following calculations: mineralizing surface/bone surface (MS/BS) = (½ sL surface + dL surface)/BS (expressed as a percent), mineral apposition rate (MAR) = IrLTh per 10 days (expressed as μm/day), and bone formation rate (BFR) = MAR × MS/BS (expressed as μm3/μm2/day) (31, 32).
Micro-CT–based angiography for studying neovascularization in the bone tunnel.
After micro-CT scanning, the other 6 rabbits were placed under anesthesia and perfused with heparinized saline. A lead-chromate, radiopaque-based contrast agent (Flow Tech) was perfused through the abdominal aorta, using our previously described protocol (21, 23, 33). Briefly, the rabbits were anesthetized by muscular injections of xylazine (20 mg/kg body weight) and ketamine (50 mg/kg body weight). After opening the abdominal cavity using an operational apparatus, the abdominal aorta and cardinal vein were separated. The blood circulation was flushed using 600 ml heparinized saline via the abdominal aorta, and then 100 ml formalin was injected. A liquid compound of MV-Diluent, MV-117 Orange, and MV Curing Agent (Microfil; Flow Tech) was injected into the abdominal aorta, and the mixture was allowed to flow through the circulation and then flushed from the cardinal vein. Thereafter, the rabbit cadavers were stored at 18°C for 4 hours.
For examination of neovascularization of the bone tunnel, the distal femora were harvested and prepared for decalcification in 9% buffered formalin acid. Micro-CT–based angiography (vivaCT, μ40; Scanco Medical) was then used to differentiate radiodensity-based segmentation of the contrast-filled vessels from the surrounding mineralized tissue (21, 23). The specimens were scanned at a 10-μm isotropic voxel size, and 2-dimensional (2-D) tomograms were reconstructed. The initial bone tunnel was defined as the volume of interest (VOI) for evaluation of the vascular network following perfusion of the vessels with contrast agent (Figures 2A and B). The vessel diameter and volume fraction within the VOI were calculated using the proper threshold of signal intensity (range 95–1,000) in the micro-CT built-in software.
Data are expressed as the mean ± SD. One-way analysis of variance was used to compare the results among fresh BMMNCs, cryopreserved BMMNCs, and steroid-treated fresh BMMNCs in the in vitro studies, and paired-sample t-tests were used to compare the results between 2 groups in the in vivo studies. P values less than 0.05 were considered statistically significant.
Activity of the stem cell pool.
The mean volume of the collected bone marrow was ∼10 ml. Thus, the mean number of BMMNCs isolated per ml of bone marrow was determined (summarized in Table 1), and the results showed that the number of fresh BMMNCs per ml of bone marrow was significantly higher before SAON induction than after SAON induction (mean ± SD 6.20 ± 0.86 × 106 versus 3.43 ± 0.83 × 106; P < 0.001).
Table 1. Number, recovery, and viability as well as colony-forming ability and adherence of fresh BMMNCs and cryopreserved BMMNCs before steroid treatment and of steroid-treated fresh BMMNCs*
Before steroid treatment
Steroid-treated fresh BMMNCs
Except where indicated otherwise, values are the mean ± SD (n = 10 per group). NA = not applicable; CFU = colony-forming units.
P < 0.001 versus steroid-treated fresh bone marrow–derived mononuclear cells (BMMNCs).
P = 0.005 versus steroid-treated fresh BMMNCs.
P = 0.006 versus steroid-treated fresh BMMNCs.
Volume of harvested bone marrow, ml
No. of BMMNCs in bone marrow, ×106 per ml of bone marrow
The total recovery of the BMMNCs and the viability of the recovered cells after freezing were evaluated. The number of Ficoll-separated BMMNCs before freezing was hypothesized to be 100%, and the recovery of these cells was ∼89.71% after cryopreservation (Table 1). The viability of the cryopreserved BMMNCs was ∼96.49%.
The number of CFU, diameter of the CFU, and number of adherent cells in cultures of fresh BMMNCs, cryopreserved BMMNCs, and steroid-treated fresh BMMNCs are also summarized in Table 1. There was no significant difference in the number of CFU and number of adherent cells between the fresh BMMNC group and cryopreserved BMMNC group. In contrast, the number of CFU in cultures of steroid-treated fresh BMMNCs was significantly lower than that in the other 2 groups (each P < 0.001). Furthermore, the number of adherent cells was significantly lower in the steroid-treated fresh BMMNC cultures when compared to the fresh BMMNC (P = 0.005) and cryopreserved BMMNC (P = 0.006) cultures before steroid treatment. There was no significant difference in the diameter of the CFU among the 3 groups.
Osteogenic differentiation of BMMNCs.
No significant difference in osteogenesis was found between the fresh BMMNCs and cryopreserved BMMNCs before steroid treatment (Table 2). The levels of the osteogenic markers AP and total calcium were, however, significantly lower in the steroid-treated fresh BMMNCs than in the pre–steroid treatment fresh BMMNCs (P = 0.002 for AP and P < 0.001 for total calcium) and cryopreserved BMMNCs (P = 0.004 for AP and P < 0.001 for total calcium) (Table 2).
Table 2. Levels of AP and calcium in fresh BMMNCs and cryopreserved BMMNCs before steroid treatment and in steroid-treated fresh BMMNCs*
Before steroid treatment
Steroid-treated fresh BMMNCs
Values are the mean ± SD (n = 10 per group). AP = alkaline phosphatase.
P = 0.002 versus steroid-treated fresh bone marrow–derived mononuclear cells (BMMNCs).
In animal experiments evaluating the bone tunnel, no rabbits died or developed inflammation at the surgical area after steroid administration. Twelve weeks after the operation, new bone had formed within the bone tunnel in both the BMMNC-treated group and the nontreated control group, as observed on 2-D and 3-D micro-CT images, with significantly more new bone formation in the BMMNC group (Figure 1, panels A2 and B2) when compared to the control group (Figure 1, panels A1 and B1). More newly formed bone within the bone tunnel in the BMMNC-treated group was also confirmed by histologic observation (Figure 1, panel C2), as compared to the more fibrous connective tissue formed within the bone tunnel in the control group (Figure 1, panel C1). When observed under a polarized light microscope, random alignment of collagen was found within the bone tunnel in the control group (Figure 1, panel D1), whereas more aligned collagen fibers were demonstrated in the BMMNC group (Figure 1, panel D2).
The histomorphometric features of the newly formed bone within the bone tunnel were significantly different between the groups (Figure 1E), as indicated by the BMD, BV/TV, Tb.N, Tb.Th, and Conn.D values, all of which were significantly higher in the BMMNC-treated group compared with the control group (P = 0.026 for BMD, P = 0.038 for BV/TV, P = 0.007 for Tb.N, P = 0.049 for Tb.Th, and P = 0.020 for Conn.D). In addition, the Tb.Sp value was significantly lower in the BMMNC-treated group compared with the control group (P = 0.001). Bone histomorphometric analyses also showed that the MS/BS, MAR, and BFR were significantly higher in the BMMNC-treated group compared with the control group (P = 0.017 for MS/BS, P = 0.027 for MAR, and P = 0.004 for BFR) (Figure 1F).
When magnified under a light microscope (at a magnification of ×20), the newly formed bone in the bone tunnel in both groups showed a normal mineralized bone matrix with lamellar patterns and lacunae filled with osteocytes (Figure 3, panels A1 and A2). However, a significant difference was found in the marrow components. In the control group, the newly formed bone was surrounded by a significant portion of fibrous connective tissue, whereas normal marrow was observed in the BMMNC-treated group (Figure 3, panels A1 and A2). Furthermore, in the control group, the subchondral bone and bone around the tunnel displayed ON lesions with empty lacunae next to the tunnel, as well as ON foci surrounded by normal bone (Figure 3, panels B1 and C1). In contrast, there were no ON lesions found next to the tunnel in the BMMNC-treated group (Figure 3, panels B2 and C2).
Neovascularization in the bone tunnel.
The 3-D microarchitecture of neovascularization in the bone tunnel was reconstructed using micro-CT–based angiography at week 12 postoperation. Newly formed vessels were observed within the bone tunnel in both the control and BMMNC-treated groups (Figures 2C and D). In addition, the voxel number and volume distribution of the newly formed vessels were assessed (Figures 2E and F). The diameter of the vessels ranged from 30 μm to 240 μm in the control group, and from 30 μm to 270 μm in the BMMNC-treated group. The peak values for the voxel number and volume fraction were concentrated in vessels with a diameter of 120 μm in the BMMNC-treated group and 60 μm in the control group. Comparison of the average volume and diameter of the neovascularization in the samples from each group, as summarized in Figure 2G, showed that the volume and diameter of the newly formed bone in the bone tunnel were both significantly higher in the BMMNC-treated group (mean ± SD 1.14 ± 0.56% and 138.77 ± 25.95 μm, respectively) compared with the control group (0.54 ± 0.38% and 115.03 ± 21.52 μm, respectively; P = 0.005 for volume fraction and P = 0.036 for diameter).
The results of the present experimental study confirm our concept of the usefulness of cryopreserved BMMNCs for enhancement of bone healing. In particular, the findings from our in vitro study showed that the bioactivity of BMMNCs is well kept after cryopreservation and is superior to the bioactivity of BMMNCs after steroid treatment. Therefore, a treatment strategy of implantation of cryopreserved BMMNCs combined with core decompression appears to be beneficial for the treatment of SAON at the early stages.
In addition, we demonstrated that cryopreservation, as a potential biotechnology for clinical applications, is an efficient method for the preservation of various types of tissue and cells, such as human skin allografts, muscle specimens, embryo bone marrow, BMSCs, and hematopoietic stem cells (26, 34–37). The present study was the first preclinical experiment to confirm that bone regeneration within a surgically constructed bone tunnel could be enhanced by the implantation of cryopreserved BMMNCs before steroid administration in a rabbit model of SAON, since cryopreservation of the cells prior to steroid administration would avoid the adverse effects of the steroids on the activity of progenitor cells in the bone marrow. The results confirmed our hypothesis that cryopreserved BMMNCs maintain their bioactivity and do in fact promote bone repair at the bone tunnel when the necrotic bone is removed via core decompression.
Results of in vitro studies by other researchers support our present findings, in that prior studies have shown that there is no difference in the osteogenic potential between cells with and cells without cryopreservation treatment, suggesting that cryopreserved BMSCs or BMMNCs may retain their bone-formation capability (38, 39). Although the current study did not compare cryopreserved BMMNCs with fresh BMMNCs in vivo, the cell viability and osteogenic differentiation potential of the cryopreserved BMMNCs in vitro were found to be similar to those of fresh BMMNCs, implying that cryopreserved BMMNCs before steroid treatment are biologically superior to steroid-treated BMMNCs for the enhancement of osteogenesis.
In assessing the relationship between the stem cell pool and repair of ON lesions, both Gangji et al and Hernigou et al have found that the decreased activity of BMSCs at the early stages of SAON results in inadequate bone repair (5–8). Steroids could induce differentiation of BMSCs into an adipocyte lineage, thereby inhibiting the osteogenic differentiation of these cells (15, 40). These findings are consistent with our experimental results in vitro, in which we found that the osteogenic differentiation capacity of the BMMNCs was decreased in rabbits with SAON when compared with rabbits without steroid treatment, i.e., the proliferative and osteogenic abilities of the cryopreserved BMMNCs were significantly higher than those of steroid-treated BMMNCs. In our in vivo study, no ON lesion was found around the tunnel and/or at the subchondral bone in the BMMNC-treated group; in contrast, we observed ON lesions with empty lacunae in the bone around the tunnel and/or at the subchondral bone in the control group without BMMNC implantation. Nevertheless, the foci of the ON lesions were found to be surrounded by normal and newly formed bone, implying that ongoing reconstructive repair of the bone is taking place. In the rabbit model of SAON used in the present study, it was technically not possible for us to collect debris directly from the ON lesions after core decompression. Therefore, we could not record the incidence of ON or the histologic features of the ON lesions.
Both osteogenesis and angiogenesis are important events during the repair of bone, especially in healing of the ON lesions (9, 41). The decreased stem cell pool activity will result in reduced osteogenic and angiogenic potential in SAON, and therefore supplementation with a sufficient amount of stem cells would be highly desirable to meet the need for local bone repair. The results of our micro-CT–based angiographic study showed that more bone formed within the bone tunnel, accompanied by more mature vasculature within the bone tunnel, in the group treated with cryopreserved BMMNCs compared to the control group.
There are no clinical reports available to indicate that the number of BMMNCs is reduced after steroid administration in patients. Nevertheless, the findings from some clinical studies have suggested that supplementation with BMMNCs is indeed beneficial for the treatment of SAON (16, 42–45). In patients with SAON, the numbers of mesenchymal stem cell (MSC) pools are decreased in the proximal femur (7). BMMNCs contain a variety of progenitor cells, such as MSCs and hematopoietic stem cells (HSCs) (46). Apart from an impaired activity of MSCs, other cell types, such as the HSC pool, have also been observed to be affected in rabbit SAON (9). In the present study, the cryopreserved BMMNCs contained the original progenitor cells, similar to those found in fresh BMMNCs in the bone marrow. This mixture of progenitor cells might therefore play a more important role than that of a single cell line during tissue repair.
MSCs could differentiate into osteoblasts and, thus, facilitate new bone formation. Neovascularization is thought to be the result of the activity of HSCs and endothelial progenitor cells, which is an important parallel event contributing to new bone formation during bone repair (9). The results from our in vivo experiment indicated that more bone and a greater extent of neovascularization were found within the bone tunnel in the BMMNC-treated group compared to the control group, suggesting that the MSCs and/or HSCs involved in cryopreserved BMMNCs might play an important role in new bone formation. This implies that the cryopreservation of autologous BMMNCs harvested before steroid administration from patients at high risk of developing ON may be a potential approach for improving the treatment of ON lesions in the early stages of SAON.
Core decompression has been proven to be an effective treatment for ON, as it can delay total joint replacement by at least 5 years, if the procedure is carried out when the ON lesion is still small. However, concerns still remain, including the fact that there is incomplete reconstructive repair and a potential for weakening of the trabecular bone within and next to the necrotic region when the ON lesion is relatively large (10–14). In addition, once the femoral head develops to a subchondral fracture, especially to the point of collapse and development of osteoarthritis, followed by narrowing of the joint space, the treatment efficacy of core decompression is significantly reduced.
Therefore, the Association Research Circulation Osseous Study Group has recommended several options for the treatment of SAON at different stages, including both medication and surgical therapy, such as core decompression combined with cell supplementation, bone grafting, bone marrow, or total arthroplasty. However, if the bone tunnel or defect is too large, cell-based tissue engineering, e.g., implantation of porous scaffold seeded with progenitor cells, may also be expected to achieve better treatment outcomes. In the current study, using our established rabbit model of SAON, we introduced a promising new treatment strategy, involving the use of cryopreserved BMMNCs for promoting bone repair within the surgically constructed bone tunnel.
The present study has some limitations. This rabbit model of SAON is a commonly used animal model with a high incidence of ON, as has been reported by our research group and by others (21–24). However, when compared with ON in humans, the major limitation or difference in this animal model is that the ON lesions do not lead to joint collapse in rabbits. This difference is attributed to the dissimilarity in weight-bearing between such quadrupedal animals and bipedal humans, especially at the hip joint. Differences in SAON risk associated with joint-loading conditions were also previously noted in a study of SAON-induced hip joint collapse in bipedal emus (47). Similarly, no changes associated with a risk of joint collapse could be found in the rabbits in our present model.
Whether our experimental findings could be directly translated into clinical practice still depends on a number of factors, including the availability of proper facilities, predictive power of the ON diagnosis, willingness of the patients to provide BMMNCs for cryopreservation, and cost effectiveness of such an approach. If we could increase the power to differentiate patients at a truly high risk of developing SAON, then we could also increase the predictive power of the diagnosis of SAON, e.g., with the development of more effective biologic tests and imaging technologies, such as analyses of lipid metabolism and clotting disorders, gene analysis in patients with ON (to identify single-nucleotide polymorphisms) (48), or functional perfusion magnetic resonance imaging (24). Development of all of these techniques would increase the potential of our concept of implantation of cryopreserved BMMNCs for the treatment of SAON in high-risk patients in clinical practice.
In the present experiments, the cryopreserved BMMNCs were indeed proven to enhance vigorous bone regeneration in the bone tunnel and around ON lesions. However, the bone tunnel created during core decompression was not a critical bone defect, which indicates that this method might only be effective when the bone defect is relatively small. The effect of cryopreserved BMMNCs on larger ON lesions, especially at weight-bearing regions, should be further examined before the strategy can be recommended for clinical applications.
In summary, our findings support the hypothesis that cryopreserved BMMNCs maintain their proliferative and osteogenic differentiation capacities, thus enabling them to promote bone repair and vascularization within a surgically constructed bone tunnel after core decompression in a rabbit model of SAON. Our experimental evidence might support a potential strategy of cryopreservation of normal autologous BMMNCs as a way to achieve good results after core decompression in high-risk patients, before long-term or pulsed administration of steroids. A joint effort or collaboration between rheumatologists and orthopedists is crucial for realizing the success of such a clinical strategy.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Drs. Zhang and Qin had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Xie, Wang, Zhang, Qin.
Acquisition of data. Xie, Wang, He, Liu, Sheng.
Analysis and interpretation of data. Xie, Wang, He.