Silicate Granules Preconditioned with Human Bone Marrow Mononuclear Cells Improve Osteogenesis in Bone Sarcoma Patients

The surgical resection of bone tumors often leads to osseous nonunions and loss of structural integrity [1]. Different strategies have been used to repair and restore skeletal defects, with good mechanical and functional properties [2-8]. The role of synthetic and bone-graft substitutes is not only to replace missing bone but also to encourage bone integration, i.e., by acting as a scaffold to guide bone growth into the graft [9]. Technological advances along with a better understanding of bone-healing biology have led to the development of various porous ceramics that are currently available to orthopedic surgeons [10]. Threedimensional (3D) porous materials are therefore preferentially used to provide a better environment for cellular attachment and proliferation, and their architecture defines the ultimate shape of new bone [11]. One of the most successful synthetic bone-grafting materials is Actifuse® (Baxter Healthcare, Newbury, UK), made up of HA-substituted porous silicon granules (2-5 mm) [7,11]. Actifuse® granules bind bone via the rapid formation of a Hydroxycarbonate Apatite (HCA) surface layer and stimulates osteogenesis via their dissolution following osteoblast adhesion [8]. Currently, there is great interest in applications of biologic stimuli to enhance bone regeneration [12,13]. Several studies evaluated the capability of mesenchymal stem cells, with different origins, to improve the efficacy of bone regenerative potential of biomaterials in both animal models [14-16] and clinical settings [17-19]. Bone regeneration by autologous cell transplantation is one of the most promising treatment [17]. Mesenchymal stem cells (MSC) combined with an osteoconductive scaffold have been shown to support bone repair for bone tissue regeneration [20,21]. The main advantage in using cells as biological stimuli is due to autogenous properties. Their application can prevent possible complications, such as immunogenic reactions and disease transmissions, and are easy to prepare. The aim of the present study was to evaluate in vitro the biocompatibility, colonization and adhesion of Bone Marrow Mononuclear Cells (BMMCs) to silicate granules. Additionally, we report our findings on the use of silicate granules mixed with BMMCs in bone sarcoma patients to improve bone reconstruction.


Introduction
The surgical resection of bone tumors often leads to osseous nonunions and loss of structural integrity [1]. Different strategies have been used to repair and restore skeletal defects, with good mechanical and functional properties [2][3][4][5][6][7][8]. The role of synthetic and bone-graft substitutes is not only to replace missing bone but also to encourage bone integration, i.e., by acting as a scaffold to guide bone growth into the graft [9]. Technological advances along with a better understanding of bone-healing biology have led to the development of various porous ceramics that are currently available to orthopedic surgeons [10]. Threedimensional (3D) porous materials are therefore preferentially used to provide a better environment for cellular attachment and proliferation, and their architecture defines the ultimate shape of new bone [11]. One of the most successful synthetic bone-grafting materials is Actifuse® (Baxter Healthcare, Newbury, UK), made up of HA-substituted porous silicon granules (2-5 mm) [7,11]. Actifuse® granules bind bone via the rapid formation of a Hydroxycarbonate Apatite (HCA) surface layer and stimulates osteogenesis via their dissolution following osteoblast adhesion [8]. Currently, there is great interest in applications of biologic stimuli to enhance bone regeneration [12,13]. Several studies evaluated the capability of mesenchymal stem cells, with different origins, to improve the efficacy of bone regenerative potential of biomaterials in both animal models [14][15][16] and clinical settings [17][18][19]. Bone regeneration by autologous cell transplantation is one of the most promising treatment [17]. Mesenchymal stem cells (MSC) combined with an osteoconductive scaffold have been shown to support bone repair for bone tissue regeneration [20,21]. The main advantage in using cells as biological stimuli is due to autogenous properties. Their application can prevent possible complications, such as immunogenic reactions and disease transmissions, and are easy to prepare. The aim of the present study was to evaluate in vitro the biocompatibility, colonization and adhesion of Bone Marrow Mononuclear Cells (BMMCs) to silicate granules. Additionally, we report our findings on the use of silicate granules mixed with BMMCs in bone sarcoma patients to improve bone reconstruction.

Biomaterials
Porous silicate substituted HA Actifuse® (0.8 wt. % Si) purchased from Baxter Healthcare, Newbury, UK) is composed by granules with sizes typically between 1 and 3 mm. It was prepared with technology based on slurry expansion (following instruction procedures). Briefly, slurry with a high powder concentration was used and expanded in a known volume to achieve a total porosity of 80% in volume, which corresponds to a large surface area (0.9 m 2 /g); its porosity is characterized by bi-modal porous structures and controlled morphology. Demineralized bovine bone matrix (DBM) particles are obtained from bovine cortical bone with sizes between 1 and 3 mm. Their porosity equals natural human bone-micropores with a diameter of 1-2 m that were frequently seen. The DBM particles are sterilized by a validated per-acetic acid-based process (data published by the manufacturer (LifeNet health Italy).

Isolation and growth of BMMSCs
Iliac crest bone marrow aspirates (20 ml) were obtained from four orthopedic patients under general anesthesia according to the ethical committee of the Istituto Nazionale Tumori G. Pascale Napoli Italy. All donors provided informed consent. Bone marrow mononuclear fraction was isolated by Ficoll-mediated (Histopaque, 10771, Sigma Company, Milan, Italy). Briefly, bone marrow aspirate was diluted in Hanks balanced salt solution (HBSS) to make the volume up to 30 ml. Cell solution was, gently overlaid on 60 ml of Histopaque. The layer at the interface of the Ficoll and HBSS was collected after 30 min of centrifugation at 1,800 g at room temperature without brake. The interface cell layer was transferred and cell suspension was centrifuged at 1,000 g for 10 min at room temperature. The pellet was suspended and cells were seeded in a T75 flask and cultured with αMEM (Lonza, Milan, Italy) supplemented with 20% fetal bovine serum (FBS), 100 units/mL penicillin (Euroclone, Wetherby, UK), 100 mg/ml streptomycin, and 1% GlutaMAX (Gibco Invitrogen, Paisley, Scotland), and incubated at 37°C with 5% CO 2 in a humidified atmosphere.

Seeding of bone marrow mononuclear cells
Plates were incubated with a mix of 1ml of slurry Actifuse® granules or demineralized bovine bone in 1 mL fibronectin solution (10 g/ mL, Sigma, Deisenhofen, Germany) in PBS without Mg 2+ and Ca 2+ (PBS−/−) for 30 min, forming a dense monolayer. The supernatants were removed and the granules were air-dried under sterile conditions at room temperature. Then, 2 × 10 4 cells/well were seeded into 24-well pre-coated plates in EGM-2 containing 0.5% Fetal Calf Serum plus growth factors (Lonza). After different times of incubation under the appropriate conditions, the medium containing the non-adhering cells was removed and rinsed once again over the bone-graft layer. This procedure was repeated three times. The (DBM and granules) were then gently transferred to another well. The remaining cells in the supernatant together with the bottom of the initial seeding well were collected and spin at 1,500 g for 5 min. Cells were stained with 0.75% crystal violet in a solution of 50% ethanol, 0.25% NaCl, and 1.75% formaldehyde. Absorbance was read at 595 nm with an ELISA reader (Infinite M200, Tecan, Mainz, Germany). The percentage of adherent cells was calculated: [(initial cell number-remaining cell number)/initial cell number] × 100%. All experiments were performed in triplicate.

MTT assay
For determination of the metabolic activity of the seeded cells the Cell Proliferation Kit I (MTT, Roche Diagnostics, Mannheim, Germany) was applied. The assay is based on the cleavage of the yellow tetrazolium salt MTT (3-[4,5-dimethylthiazol-2-yl]-2,5diphenyltetrazoliumbromide) to purple formazan crystals by metabolic active cells. BMMCs were seeded on a bone-graft substitutes monolayer covering in 24-well plate as described above. Before the addition of the MTT reagent the granules were transferred to an empty well in order to prevent false positive results caused by cells adhering to the bottom of the well. Ninety μl medium and 10 μl of MTT labelling reagent were added to each well and cells were incubated additional 4 h. Next, the cells were incubated overnight with a solubilization solution. The supernatants were collected and transferred to another 96-well plate. Then the absorbance at 570 nm was measured with an ELISA reader (Ceres UV900c, Bio-Tek Instruments, Windoski, VT, USA). MTT levels are normalised to the number of cells in the scaffolds. As control increasing numbers (1000, 2500, 5000, 10,000) of BMMCs, directly seeded in 96-well plate, were assessed as well.

Scanning Electron Microscopy (SEM) analysis
For SEM analysis, BMMCs (10 4 cells/well) were cultured on silicate granules or DMB for 72 h and processed as described previously. Briefly, cells attached to the biomaterials were washed with PBS, fixed in 4% paraformadehyde/PBS and dehydrated with increasing ethanol percentage (30-90% in water for 5 minutes and twice with 100% ethanol for 15 minutes) then treated in Critical Point Dryer (EMITECH K850) and finally sputter coated with platinum-palladium (Denton Vacuum DESKV) and observed with Supra 40 FE-SEM (Zeiss).

Patients
Twenty patients with osteolytic malignant lesions in different bone segments underwent to clinical, radiological, histopathology evaluation and following surgical resection. All data was recorded on a prefixed proforma (Table 1 and Supplementary Table 1). The lesions were located in 12 patients at the proximal femoral metaphysis, in 4 patients at the proximal tibia, in 3 patients at the proximal humerus and 1 patient at the knee. The histological diagnoses of lesions were (n=10) GCT (Giant cell tumor), (n=4) aneurysmal bone cyst, (n=3) epiphyseal chondroblastoma and (n=3) chondromyxoid fibroma. The mean volume of the lesion, measured with a preoperative computational tomography (CT), was 18.5 cm 2 (range [16][17][18][19][20][21][22][23][24]. After surgery all patients were subjected to clinical and radiograph examination every 2 weeks during the distraction phase and every month thereafter until the end of the consolidation phase (one year). Follow-up evaluation was done by using the V.A.S. scale [22,23] and Musculoskeletal Tumour Society (MTS) scoring system [24]. Numerical values from 0 to 5 points were assigned for each of the following 6 categories: pain, function, emotional acceptance, use of supports, walking ability and gait. These values were added, and the functional score was presented as a percentage of the maximum possible score. The results were graded according to the following scale: Excellent-75% to 100%; good-70% to 74%; moderate-60% to 69%; fair-50% to 59% and poor-<50%. This study was approved by the institutional review board of the author's institution Istituto Nazionale Tumori G Pascale Napoli was performed in accordance with the ethical standards of the 1964 Declaration of Helsinki as revised in 2000. The bone reconstitution stage was determined by clinical and radiograph examination every 2 weeks during the distraction phase and the callus type was classified according to the International system [25].

Surgery and preconditioned silicate granules
Patients were subjected to surgery according to international guidelines under general anesthesia. In the first part of the surgery 20 ml of autologous bone marrow was aspirated with "Jamshidi" 8-gauge syringe, containing anticoagulant citrate dextrose, from the anterior superior iliac spine. After incision and excision of the biopsy tissue, it was opened a wide bone window through which it was possible to make careful a curettage. Extended curettage was done using a high speed (7,000 g, Midasrex©) Phenol (1ml of melted phenol mixed with 10ml of normal saline) as chemical adjuvant in all cases. The cavity was then packed with autologous BMMCs and/or silicate granules (ranging from 10 ml to 40 ml). The mean ratio between BMMCs and silicate granules was (0.3-1 × 10 7 BMMCs /10 ml granules Actifuse®).

Statistical analysis
Data are expressed as the mean of ± standard deviation. Statistical analyses of cell-biology experiments, which were performed in triplicate, were carried with T student-test analysis (P<0.05 was considered significant).

Results
To monitor the capability and time of BMMCs to seed on granules, mononuclear fraction was plated on dishes coated with silicate granules or with DBM (see methods). The number of cells present in medium, following 3-time washes, were monitored at different time points during the assay period and reported as percentage of seeding cells (see methods). As shown in Figures 1A, in 30 minutes 90% of cells adhered to silicate granules showing an exponential progression. In contrast 20% of BMMCs seeded on DBM after 30 minutes and 5% in matrigel coated dishes respectively (P<0.001 and P<0.05). Survival of cells were observed up to day 10 by the MTT assay on each scaffold with active metabolism. No differences were observed in all cases as showed in Figures 1B. Moreover, BMMCs seeded in plate at low density (1,000 cells/cm 2 ) gave a spherical phenotype, often accumulated in clusters and were capable to form colonies as shown in Figures 1C. In order to evaluate the fractions of cells with putative regenerative capacities, cells following seeded on silicate granules, were analyzed by flow cytometry in agreement with the International Society for Cell Therapy guidelines [26]. In the analyzed samples, 85% of the cells co-expressed CD105   Figure 2.
Scanning electron microscopy showed that BMMCs grown on silicate granules had normal morphology and appeared to be well attached to the substrate with long pseudopodia in contact with the extra-cellular matrix (Figures 3A and 3B). The amount of debris on the cell surface was low, according to this type of biomaterials. Different behavior was registered towards DBM attached cells showed a low number of shorter pseudopodia and greater amount of cells in natural bone niche ( Figures  3C and 3D). For in vivo study, we analyzed twenty bone sarcoma patients whose the clinical pathological characteristics are reported in Table 1 and supplementary Table 1. At the clinical examination, the patients presented spontaneous pain localized to the site of the lesion with functional analgesic limitation. The pain was exacerbated by acupressure and it was possible to appreciate the tumefaction of the tumor. The mean age of patients was 27 years, tumors histology were different and localized in different sites as indicated in Table 1 and  supplementary Table 1. The radiological examination, before surgical resection of tumor showed the presence of a large osteolytic area with mean of 18.5 cm 2 . A representative cases are showed in Figure 4A and 4D. A total of 16 patients underwent to reconstructive curettage and filling of the lesion site with silicate granules alone. The Figure 4A reported the case of patient affect from chondroblastoma of knee filled with silicate granules alone. After 15 days, the patient showed a callus shape at X-rays examination ( Figure 4B and 4C) no complication were reported (Table 2) and the functional outcome following Musculoskeletal Tumor Society score (MSTS) was of 60% (Tables 2 and 3). The Figure 4D reported a representative patient subjected to curettage and filling of humerus lesion with silicate granules combined with BMMCs. The X-rays examination of bone cavity revealed the presence of callus after two weeks ( Figure 4E and 4F). Silicate granules introduced into the humerus cavity showed low material around the periphery of the pores indicated bone formation within the marrow cavity. The outcome at two weeks of humerus filled with silicate granules and BMMCs was 78% by MSTS score compared to 58% of humerus filled with silicate granules alone (Table 3 and Supplementary Table 1). After 4 weeks all patients treated had bone cavities completely occupied by bone callus as reported in Tables 2 and 3. None patients showed necrosis or infection of lesions ( Table 2) that required additional surgery of bone.

Clinical outcome
According to the V.A.S. evaluation system there were no significant differences at 2, 4, 48, weeks, after the surgery. According to the MSTS evaluation system, after the surgery, we found that patients treated with curettage and filling of the defects with silicate granules and BMMCs had mean of MSTS score of 82% (Table 2 and Table 3). In contrast patients treated silicate granules alone the mean of MSTS score was 60%. During the follow up (48 weeks) none patients develop infection or hematoma that requires surgical revision neither necrosis. The encouraging preliminary results are reported in Tables 2 and 3.

Discussion
Research in tissue engineering is focused on finding new approaches for bone regeneration that optimize results by seeding BMMCs on a porous ceramic scaffold. Several orthopedic studies in large animals emphasize the difference between the healing of sites treated with scaffolds loaded with BMMCs compared to scaffolds without BMMCs [27,28]. Moreover, silicate granules has shown promising results, in bone regenerative therapy with fast bone apposition rates [28]. In human, ex vivo studies have investigated biocompatibility between stem cells and synthetic material for bone regeneration [29]. In patients progress has been made in ligament regeneration using BMMCs [30] and several trials are ongoing [31]. Our preliminary study utilizes, for the first time, a combination of silicate granules with BMMCs to improve bone repair and to provide immediate support for large tumor cavities in human [32]. To characterize the behavior in vitro of silicate granules with bone marrow aspirates from orthopedic patients we cultured BMMCs in close contact with selected granules. Viability assays demonstrated that the biomaterial did not elicit any cytotoxic effects. The positive charge present on the silicate material promoted cell adhesion in a few minutes [29]. Additionally, fluocytometry analysis indicated that most of adhered cells were mesenchymal with high regenerative potential. Findings from SEM analysis showed a well-organized cytoskeleton architecture and long cytoplasmic bridges between cells and substrate. The ability of BMMCs to contact a large area of the silicate material is an important indicator to evaluate long-term healing and stability of silicate granules in human bone. Our preliminary findings demonstrated that the combinatory use of autologous BMMCs and silicate granules improved bone regeneration within lesion cavity detected as formation of callus and a mean MSTS of 85% at two week's follow-up. Further studies will be necessary to determine why and under which conditions the new bone wall develops and new cartilage grows on regenerated bone.

Conclusions
In this preliminary study, we investigated bone reconstruction using silicate granules in combination with autologous BMMCs. Patients with a small and medium-sized bone cavity (mean 18.5 cm 2 ) had successful single-stage outcomes. In summary, post-surgery use of silicate granules in combination with autologous BMMCs to fill the bone lesion appears to improve rapidity of bone integrity reconstruction. However, a larger study cohort and longer follow-up times are required to identify additional predictors and indications.