A comparison of bone formation in biphasic calcium phosphate (BCP) and hydroxyapatite (HA) implanted in muscle and bone of dogs at different time periods

Physicochemical modification could implement synthetic materials into osteoinductive materials, which induce bone formation in nonosseous tissues. We hereby studied the relevance between the osteogenic capacities of osteoinductive materials in nonosseous tissues and in osseous sites. Biphasic calcium phosphate ceramic (BCP) and hydroxyapatite ceramic (HA) were implanted in femoral muscles and femoral cortical bone of dogs for 7, 14, 21, 30, 45, 60, 90, 180, and 360 days, respectively. Two dogs were used in each time point. In each dog, four cylinders (phi5x6 mm) per material were implanted in femoral muscles and 2 cylinders (phi5x6 mm) per material in femoral cortical bone. The harvested samples were processed for both histological and histomorphometric analyses. Bone was observed in BCP implanted in femoral muscles since day 30, while in HA since day 45. Quantitatively, more bone was formed in BCP than in HA at each time point after day 30 (p<0.05). The earlier and more bone formed in BCP than in HA suggests BCP a higher osteoinductive potential than HA in muscle. In femoral cortical bone defects, a bridge of bone in the defect with BCP was observed at day 21, while with HA at day 30. At days 14, 21, and 30, significantly more bone was formed in BCP than in HA (p<0.05). The results herein show that osteogenic capacities of osteoinductive materials in nonosseous tissues and osseous sites are correlated: the higher the osteoinductive potential of the material, the faster the bone repair.     

Histomorphological researches on large porous hydroxyapatite cylinder tubes with polylactic acid surface coating in different nonskeletal sites in vivo

Porous hydroxyapatite (HA) ceramic cylinder tubes coated with polylactic acid on the exposed surfaces were implanted in four nonskeletal sites (omentum, peritoneum, vastus lateralis, and side of femur). Six months postoperatively, proper amount of Chinese ink was injected to dye the implanting areas. Decalcified and nondecalcified sections were observed under inverted microscope. The results showed that the soft tissues around the HA cylinder tubes in peritoneum, vastus lateralis, and side of femur groups appeared visible black. Some small blacked vascular architectures were also discernible. However in omentum group, only small number of blacked vessels existed. Histological observations indicated that vascularization and ossification occurred in peritoneum, vastus lateralis, and side of femur groups. In omentum group, there was no any sign of vascularization and ossification. A conclusion could be made in this study that excepting bones and muscles, parietal peritoneum could serve as a potential spot for culturing histoengineering hydroxyapatite (HA)-polylactic acid (PLA) ceramic bone substitutes.     

Replacement of segmental bone defects using porous bioceramic cylinders: a biomechanical and X-ray diffraction study

A porous ceramic material [hydroxyapatitetricalcium phosphate (HA-TCP)] was implanted in the femora of 30 dogs to investigate the possibility of using this material to repair segmental bone defects. A bone segment, 1.5 cm in length, was removed from the diaphysis of one femur in each dog to create the defect. Cylinders of corresponding size were inserted into the defects. The animals were divided into three groups with recovery times of 2 months, 4 months, and 6 months, respectively. The implants were harvested and subjected to biomechanic tests (bending strength) and X-ray diffraction analysis. The bending strengths of the implant construct increased gradually over time postoperatively. The values of strength for the three different time groups had significant variations (p < 0.05). The X-ray diffraction analysis indicated that the peaks of the TCP included in the cylinders decreased in intensity after implantation and tended to be similar to those of natural bone by 6 months after operation. Conversely, the peaks for the HA had fewer changes compared with preimplantation values. Based on the results of this experiment it was concluded that the porous HA-TCP ceramic cylinders have potential for repair of segmental bone defects if assisted by adequate stabilizing fixtures during the early postoperative period.     

Osteoinductivity and biomechanics of a porous ceramic with autogenic periosteum

Twenty-one dogs were used to study the osteoinductivity and biomechanical properties of a biphasic porous ceramic with autogenic periosteum implanted in muscle. The ceramic implants were swathed in fresh periosteum derived from the same animals and implanted in the femur muscles. The other two groups of animals served as controls using the same material implanted in the femur bones and muscles without periosteum. Biomechanical measurements showed that, in the muscles, the experimental group had a higher bending strength than the unswathed group by the time the samples were harvested. Six months postoperatively, the strength of the samples in the experimental group had almost reached that of normal bones. The results of X-ray diffraction and infrared spectrometric analysis suggested that the degradation rate and speed of tricalcium phosphate (TCP) of the ceramic in the experimental group were faster than in the unswathed samples, but slower than in samples implanted in bones. The bone replacement and bone-inducing activity were excellent in the periosteum-swathed samples. Histologically, satisfactory bone repair was seen in the experimental samples. All results indicate that autogenic periosteum could increase bioactivity of ceramics in heterosites and improve bone formation in the surroundings of porous calcium phosphate ceramics. The data also infer that the complicated procedure of culturing bone growth factors with biomaterials in vitro to obtain bioactive grafts could be replaced by this relatively simple method.     

Osteogenicity of biphasic calcium phosphate ceramics and bone autograft in a goat model

The aim of this work was to compare the osteogenicity of calcium phosphate ceramic granules with autologous bone graft in ectopic and orthotopic sites. Biphasic calcium phosphate (BCP) granules composed of hydroxyapatite (HA) and beta-tricalcium phosphate (beta-TCP) in a 60/40 ratio were sintered at 1050, 1125 and 1200 degrees C, producing different microporosities. Either BCP ceramic granules or autologous bone chips (n=7) were implanted into paraspinal muscles. Osteoinduction was not observed in either the BCP implants or autologous bone chips after 6 or 12 weeks in the ectopic sites. Hollow and bored polytetrafluoroethylene (PTFE) cylinders were filled with autologous bone, BCP granules or left empty, then implanted into critical-sized defects in femoral epiphyses. The PTFE cylinders left empty contained marrow and blood vessels but not mineralized bone, indicating that this model prevented bone ingrowth (0.56+/-0.43% at 12 weeks). Bone formation was observed in contact with the BCP1050 and BCP1125 granules in the femoral sites after 6 weeks. The amount of bone after 12 weeks was 5.6+/-7.3 and 9.6+/-6.6% for BCP1050 and BCP1125, respectively. Very little bone formation was observed with the BCP1200 implants (1.5+/-1.3% at 12 weeks). In both the ectopic and orthotopic sites, autologous bone chips were drastically resorbed (from 19.4+/-3.7% initially to 1.7+/-1.2% at 12 weeks). This study shows that synthetic bone substitutes may have superior stability and osteogenic properties than autologous bone grafts in critical-sized bone defects.     

骨诱导性磷酸钙陶瓷为支架的组织工程骨修复狗下颌骨箱状缺损*☆◆

背景：应用不外加生长因子或细胞而具有骨诱导性的生物材料,在非骨部位构建骨移植物,即体内组织工程骨,其在修复箱状及节段性骨缺损方面,具有更可行的前景。 目的：采用骨诱导性钙磷陶瓷材料构建体内组织工程化类骨移植物,探索其应用于修复实验动物下颌骨箱状骨缺损的可行性。 方法：以骨诱导性磷酸钙陶瓷材料为支架植入狗肌肉内构建体内组织工程骨,同期在狗自体下颌骨左右两侧各拔除牙弓中段牙2颗,形成约20 mm无牙区。8周后在无牙区形成箱状缺损,同期取出支架即刻移植入一侧自体下颌骨缺损区,对侧骨缺损区直接移植入未经体内构建的磷酸钙陶瓷作为对照。 结果与结论：经肌肉内构建的体内组织工程骨移植物的力学性能较单纯磷酸钙陶瓷有明显提高。颌骨缺损区的核吸收强度明显强于对照区,其移植物内长入的骨组织较多,两者的成骨面积差异有非常显著性意义(P < 0.01)。说明在修复颌骨大范围缺损中,体内组织工程骨移植物较单纯骨磷酸钙陶瓷替代材料表现出明显的力学和生物学优势,修复效果显著,有良好的应用前景。关键词：骨诱导性磷酸钙陶瓷;体内组织工程骨;骨缺损;支架;组织工程 缩略语注释：BCP：biphasic calcium phosphate,双相磷酸钙 doi:10.3969/j.issn.1673-8225.2012.16.001     

不同长度兔桡骨缺损区骨形态发生蛋白2的表达

背景：有研究表明在骨缺损的状态下,骨形态发生蛋白2的分布存在一定的规律。 目的：观察不同长度兔桡骨骨缺损部位骨形态发生蛋白2的表达。 方法：将48只新西兰大白兔随机分成2组,分别在麻醉下用线锯造成兔左前臂桡骨中段0.5和3.0 cm的骨缺损。 结果与结论：Western blot检测显示,0.5 cm骨缺损组骨缺损部位骨形态发生蛋白2蛋白表达量在损伤后1,3,4周呈逐渐增多的趋势,且每组均较损伤即刻明显增加(P < 0.05);3.0 cm骨缺损组骨缺损部位的骨形态发生蛋白2的相对表达量在损伤后3周达到高峰(P < 0.05),其峰值水平明显高于0.5 cm骨缺损组(P < 0.05),基本维持在较高水平。对骨痂的生成情况进行量化评估结果显示,损伤后3,4周,3.0 cm骨缺损组的骨痂生成量较0.5 cm骨缺损组明显减少(P < 0.05)。证实,兔桡骨3.0 cm的骨缺损部位骨形态发生蛋白2表达明显上调,但还不能使该长度的骨缺损自行愈合。     

Effects of FGF-2 release from a hydrogel polymer on bone mass and microarchitecture

Bone substitutes are widely used for filling and restoring bone defects. Among them, methacrylic polymers are employed in load-bearing bones to seal hip prostheses. Incorporation of growth factors into a polymer device could be a way to enhance bone growth. In the present study, we evaluated the capacity of poly(2-hydroxyethyl methacrylate) - pHEMA - copolymerized with 2-vinyl pyrrolidone - VP - to release proteins. Fibroblast growth factor-2 (FGF-2) was incorporated into cylinders of p(HEMA-co-VP). FGF-2 release was studied by ELISA in vitro and cylinders were implanted in the femoral condyle of white New Zealand rabbits. After 2 months post-surgery, FGF-2 was able to enhance bone formation by increasing bone volume; this effect was evidenced by an increase in trabecular number and bone gain was mainly in the form of woven bone. At 3 months post-surgery, no difference could be evidenced between animals receiving vehicle or FGF-2. Animals receiving vehicle exhibited bone mass higher than at 2 months and woven bone was replaced by mature bone with a lamellar matrix. The hydrogel polymer allowed the release of FGF-2, which in return enhanced bone regeneration soon after surgery but the effect vanished rapidly.     

A preliminary study on osteoinduction of two kinds of calcium phosphate ceramics.

With respect to the effect of material factors on calcium phosphate biomaterial-induced osteogenesis, the osteoinductive property of two kinds of porous hydroxyapatite ceramics, which were made by different producers, was investigated in dorsal muscles of dogs. One hydroxyapatite ceramic (S-HA), macroporous implants with rough pore walls containing abundant micropores, was made by Sichuan Union University (Chengdu, China); the other hydroxyapatite ceramic (J-HA), porous implants with smooth macropore walls composed of regularly aligned crystal grains, was provided by Mitsubishi Ceramic Int. (Japan). Different tissue response was detected histologically and microradiographically after the ceramic samples had been implanted in dorsal muscles of dogs for 3 and 6 months. Bone formation was found in S-HA at 3 months, which increased at 6 months. In contrast, no bone formation was detected in J-HA at both 3 and 6 months. These results indicate that with the special architecture, calcium phosphate ceramic can induce bone formation in soft tissue. As both materials were very similar in their chemical and crystallographic structures, but varied in their microstructures, the latter seem to be an important factor affecting the osteoinductive capacity of calcium phosphate ceramics. These data suggest that, by controlling the preparation of calcium phosphate ceramic, bone substitutes with intrinsic osteoinductive property can be developed from calcium phosphates.     

In vivo study on biocompatibility and bonding strength of hydroxyapatite-20vol%Ti composite with bone tissues in the rabbit

Biocompatibility and bonding strength of hydroxyapatite-20vol%Ti composite fabricated by hot-pressing technique with bone tissues in the rabbit were investigated by in vivo studies in comparison with those of Ti metal and dense HA ceramic. Although fibrous tissues formed at the interface between the composite and bone tissues at 3 weeks in vivo, bonding strength of the composite increases faster than that of dense HA after 4 weeks. At 3 months in vivo, bonding strength of the composite is higher than that of dense HA and exceeds 6.5 MPa. Moreover, as compared with the visible bonding interfaces between dense HA and new bones, the bonding interfaces for the composite cannot already be distinguished and the composite was osseointegrated fully with bone tissues into one bony body. The shear fracture of bonding strength test for the composite occurred in new bone zones near the interface, which indicates that bonding strength of the composite could even exceed the shear strength of new bones after 3 months in vivo. In conclusion, HA-Ti composite has better osteoconduction and osseointegration abilities than Ti metal and dense HA ceramic after 3 months in vivo and is a promising biomaterial for hard tissue replacement.     

Bone induction by porous glass ceramic made from Bioglass (45S5)

Porous glass ceramic, which was prepared from Bioglass powder (45S5, U.S. Biomaterials) by foaming with diluted H(2)O(2) solution and sintering at 1000 degrees C for 2 h, was implanted as cylinders (5 mm in diameter and 6 mm in length) in thigh muscles of dogs for 3 months. Histological observation was made on thin un-decalcified sections. Bone formation was histologically found in pores of all implants (X16) retrieved from 16 dogs. The bone tissue was also identified with backscattered scanning electron microscopy observation (BSE) and energy dispersive X-ray microanalysis (EDX). This is the first report of bone induction in soft tissues of animals by glass ceramic that has long been recognized as a bioactive (osteoconductive) biomaterial. The present results justify the impetus to investigate the osteoinductivity of calcium phosphate-based biomaterials, to study the mechanism of bone induction (osteoinduction) by calcium phosphate-based biomaterials, to develop osteoinductive calcium phosphate-based biomaterials, and to examine the relation between osteoinduction and osteoconduction.     

管状和柱状磷酸三钙陶瓷人工骨修复长骨大段骨缺损的对比研究

为研制理想的、能较快修复长骨大段骨缺损的陶瓷人工骨 ,我们将管状磷酸三钙 ( TTCP)和柱状磷酸三钙 ( CTCP)陶瓷人工骨分别植入兔桡骨于 1cm缺损处。术后 4、12周时取材 ,作大体组织形态学、新骨形成定量观察及抗折强度测试。术后 4、12周时 TTCP内新骨形成量明显多于 CTCP,术后 12周时 ,植入材料与宿主骨结合紧密 ,TTCP的降解速度快于 CTCP,抗折强度前者明显高于后者 ( P<0 .0 1)。研究结果表明 ,TTCP比 CTCP设计更合理 ,能更快促进长骨大段骨缺损的修复 ,是一种较理想的人工骨材料     

Small-animal models for testing macroporous ceramic bone substitutes

The aim of this study was to compare the bone colonization of a macroporous biphasic calcium phosphate (MBCP) ceramic in different sites (femur, tibia, and calvaria) in two animal species (rats and rabbits). A critical size defect model was used in all cases with implantation for 21 days. Bone colonization in the empty and MBCP-filled defects was measured with the use of backscattered electron microscopy (BSEM). In the empty cavities, bone healing remained on the edges, and did not bridge the critical size defects. Bone growth was observed in all the implantation sites in rats (approximately 13.6-36.6% of the total defect area, with ceramic ranging from 46.1 to 51.9%). The bone colonization appeared statistically higher in the femur of rabbits (48.5%) than in the tibia (12.6%) and calvaria (22.9%) sites. This slightly higher degree of bone healing was related to differences in the bone architecture of the implantation sites. Concerning the comparison between animal species, bone colonization appeared greater in rabbits than in rats for the femoral site (48.5% vs. 29.6%). For the other two sites (the tibia and calvaria), there was no statistically significant difference. The increased bone ingrowth observed in rabbit femurs might be due to the large bone surface area in contact with the MBCP ceramics. The femoral epiphysis of rabbits is therefore a favorable model for testing the bone-bonding capacity of materials, but a comparison with other implantation sites is subject to bias. This study shows that well-conducted and fully validated models with the use of small animals are essential in the development of new bone substitutes.     

Resorption of apatite-wollastonite containing glass-ceramic and beta-tricalcium phosphate in vivo

Apatite-wollastonite containing glass ceramic is considered to be difficult to resorb, but we experienced the disappearance of the porous type of Apatite-wollastonite glass ceramic particles . In this study, the resorption of porous apatite-wollastonite glass-ceramic implanted in the femurs of rabbits was investigated, and the process was compared with beta-tricalcium phosphate, a resorbable ceramics. Porous apatite-wollastonite glass-ceramic (70, 80, and 90% porosity) and beta-tricalcium phosphate (75% porosity) were implanted in the femurs of Japanese white rabbits. Samples were harvested and examined 0, 4, 8, 12, 24 and 36 weeks after implantation. Quantitative analysis of the radiographic and histologic findings was performed with NIH Image software. Radiographic examination demonstrated that the radiopacity and size of the porous apatite-wollastonite glassceramic cylinders decreased gradually after implantation. Histologic examination revealed that the surface area of the apatite-wollastonite glass-ceramic cylinders decreased continuously, and approached 20% of the original area 36 weeks after implantation. However, the resorption rate of porous apatite-wollastonite glass-ceramic was slower than that of beta-tricalcium phosphate. Toluidine blue staining showed abundant new bone formation on the surface of the apatite-wollastonite glassceramic matrix. Considering its mechanical strength, gradual resorption characteristics, and good osteochonductive activity, porous apatite-wollastonite glass-ceramic appears to be a suitable artificial bone substitutes.     

重组人骨形态发生蛋白2修饰的组织工程化骨修复眼眶骨折的实验研究

目的 探讨组织工程化骨修复眼眶骨折缺损的治疗效果.方法 体外构建以自体骨髓基质干细胞(BMSC)为种子细胞、可降解吸收的生物材料聚乳酸羟基乙酸聚合物(PLGA)为载体、重组人骨形态发生蛋白2(rhBMP-2)为生长因子的组织工程化骨,将实验动物分为对照组(植入PLGA/rhBMP-2复合物)和实验组(植入组织工程化骨),观察术后1个月、3个月和6个月伤口愈合情况、并发症及眼眶外观、CT影像学和组织学变化.结果 术后所有动物伤口愈合良好,无并发症和眼球凹陷.CT三维成像显示术后3个月实验组的缺损范围[(25.1±6.8) mm2]小于对照组[(55.3±7.7)mm2];术后6个月,实验组的眼眶骨折缺损消失,而对照组仍存在.组织学结果显示,术后1个月即可观察到实验组植入区边缘植入物开始缓慢吸收,少量成骨细胞沿支架长人材料内,而对照组未观察到;术后3个月可见实验组形成条带状新生骨长入将其分割包绕呈交叉排列,材料降解吸收明显高于对照组;术后6个月实验组植人材料完全被降解吸收,同时被新生骨组织取代,植入物与自身骨组织紧密结合,融为一体.而对照组仅部分降解吸收.结论 重组人骨形态发生蛋白2修饰的组织工程化骨具有较强的传导成骨和诱导成骨活性,生物相容性好,材料可完全降解,为骨组织取代,对眼眶骨折缺损具有较好的修复效果.     

Preparation of calcium aluminate cement for hard tissue repair: effects of lithium fluoride and maleic acid on setting behavior,compressive strength,and biocompatibility

Surface charges of electrically polarized hydroxyapatite (HAp) ceramics were demonstrated to enhance osteoconductivity in the wide gap of the canine bone whereas the bone formation processes varied according to the charge polarity. Cell reactions in the vicinity of the charged surfaces of HAp ceramics were not phagocytic absorption but rather bone formation by stimulated osteoblasts surrounding newly formed bone 7 days after implantation. The bone formation in direct contact with negatively charged ceramic surfaces suggests that the negative charges enhanced the osteobonding ability of the HAp ceramics. In the vicinity of the positively charged HAp surface, the formation of bones derived from osteoid tissues entirely occupied the 0.2-0.3-mm gaps between the HAp and the bone at 7 days. Surface charges induced by electrical polarization significantly cooperated with the innate bioactivity of HAp and reconstructed the wide bone defects more promptly than did the nonpolarized HAp ceramic surface.     

组织工程化骨修复犬牙槽骨缺损的组织学观察

目的观察组织工程化骨修复犬牙槽骨缺损过程中的新骨形成及其矿化程度。方法全麻及无菌条件下抽取犬胸骨骨髓,体外诱导培养犬自体骨髓间充质干细胞。取成年杂种犬8只,随机分成实验组和对照组,每组4只。实验组将已分化的自体成骨细胞与Bio-Oss骨胶原复合修复犬牙槽骨缺损:对照组牙槽骨缺损处仅植入Bio-Oss骨胶原。每组分别于术后4周,8周各处死2只动物,标本常规切片后行HE和改良Masson三色染色,光学显微镜下观察各组标本的组织学表现。结果实验组4周时即可见骨缺损修复区骨胶原内蓝色的新骨形成,随着时间的推移,形成新骨逐渐矿化成为红色的成熟骨组织,8周时骨缺损修复区可见大片新骨呈岛状或条索状排列,新生牙槽骨骨化效果明显优于对照组,Masson染色为红色。结论组织工程化骨促进犬牙槽骨缺损组织的修复。     

Mineralization and calcium fixation within a porous apatitic ceramic material after implantation in the femur of rabbits

Cylinders (0.8 cm long, 1.0 cm of diameter and with an axial hole), constituted, after firing, of a ceramic mixture of hydroxylapatite (HA) and beta-tricalciumphosphate (beta-TCP) in a 10:1 ratio, were implanted into mid-diaphyseal defects of one femur of 20 rabbits and stabilized with intramedullary rods. The implantation sites were checked radiographically every month, and after 3 months (3 animals) and 6 months (17 animals) the rabbits were sacrificed and the implants with the surrounding tissue were embedded in methylmethacrylate, cut to thick sections and analyzed by scanning electron microscopy (SEM). Porosimetric and x-rays diffraction analyses were carried out before and after implantation of the cylinders, and the state of mineralization at the bone-implant interface was determined by EDAX microprobe analysis. Bony callus formation started at 1 month at the osteotomy sites, as judged by radiography, but after 3 months a not-mineralized zone had still been demonstrated between bone and the implants. At 6 months, 13 implants showed themselves firmly fixed in their implantations beds, while 4 implants were only incorporated at their proximal ends. In bone contact zones, an enrichment of Ca2+ was displayed by microanalytical techniques in the outer zone of the implanted samples which may be explained by an apparent additional phase transformation of HA into TCP, thanks to the change of the Ca/P ratio, that takes place in vivo.     

Tissue-engineered bone repair of goat-femur defects with osteogenically induced bone marrow stromal cells

Tissue engineering can generate bone tissue and has been shown to provide a better means of repairing weight-bearing bone defect. Previous studies, however, have heretofore been limited to the use of nonosteogenically induced bone marrow stromal cells (BMSCs) or the application of slow-degradation scaffolds. In this study, weight-bearing bone was engineered using osteogenically induced BMSCs. In addition, coral was used as a scaffold material, due to its proper degradation rate for the engineering and repair of a goat femur defect. A 25 mm long defect was created at the middle of the right femur in each of 10 goats. The rates of defect repair were compared in an experimental group of ten goats receiving implants containing osteogenically induced BMSCs and in the control group of goats (n = 10) receiving just coral cylinders. In the experimental group, bony union was observed by radiographic and gross view at 4 months, and engineered bone was further remodeled into newly formed cortexed bone at 8 months. There was increased gray density of radiographic rays in the repaired area, which was significantly different (p < 0.05) from that of the control group. H&E staining demonstrated that trabecular bone was formed at 4 months. Moreover, irregular osteon was observed at 8 months. Most importantly, the tissue-engineered bone segment revealed a similarity to the left-side normal femur in terms of bend load strength and bend rigidity, showing no significant difference (p > 0.05). In contrast, the coral cylinders of the control group showed no bone formation. Furthermore, almost complete resorption of the carrier had occurred, being evident at 2 months in the control group. H&E staining demonstrated that a small amount of residual coral particle was surrounded by fibrous tissue at 4 months whereas the residues disappeared at 8 months. Based on these results, we conclude that engineered bone from osteogenically induced BMSCs and coral can ideally heal critical-sized segmental bone defects in the weight-bearing area of goats.     

Long-term stability of bone tissues induced by an osteoinductive biomaterial, recombinant human bone morphogenetic protein-2 and a biodegradable carrier

The long-term stability of bone tissues induced by recombinant human bone morphogenetic protein-2 (rhBMP-2) and poly[L-lactide-co-glycolide] copolymer-coated gelatin sponge (PGS) was examined. In 16 dogs, 2.5 cm unilateral bone defects were created in the left tibial diaphyses. Tibia was fixed with metal plate, and PGS impregnated with (0.4 mg/cm(3)) or without rhBMP-2 was implanted into 15 or one defects, respectively. The metal plates of rhBMP-2-treated limbs were removed 16 weeks after the implantation. The bilateral tibiae of five animals each of the rhBMP-2-treated group were harvested at 32, 52 or 104 weeks, and served for biomechanical testing and histology. Although the defect that received PGS alone resulted in nonunion at 16 weeks, all defects treated with rhBMP-2 achieved radiographic bony union by 8 weeks. Biomechanical properties of the regenerated bones restored to the levels of intact tibiae at 32 weeks, but torsional stiffness was significantly higher. No statistical significances were detected in all parameters between regenerated and intact tibiae at 104 weeks. No radiographic and histological findings suggesting enhanced resorption to the regenerated bones were observed. These results suggest the long-term stability of the bone tissues induced by rhBMP-2, and the usefulness of rhBMP-2-impregnated PGS as a biomaterial for long bone defect filling.