| Native bovine bone morphogenetic protein in the healing of segmental long bone defects | ||
|---|---|---|
| Prev | Chapter 2. Review of the literature | Next |
Segmental long bone defects have been used as models for bone reconstruction to evaluate different transplant materials as well as the efficacy of BMP. This model is valid in studying osteoconductive agents when the defect (large enough) does not heal spontaneously (Einhorn 1999). Animal studies with bone defects treated with bone substitute materials or BMP include dog radius (Johnson et al. 1996a, Johnson et al. 1996b, Heckman et al. 1999, Sciadini & Johnson 2000), dog femur (Johnson et al. 1996a, Bruder et al. 1988), dog fibula (Enneking et al. 1975, Burchardt et al. 1978), sheep tibia (Marcacci et al. 1999), rabbit ulna (Bolander et al. 1986, Hopp et al. 1989), rabbit radius (Zellin & Linde 1997, Teixeira & Urist 1998, Wheeler et al. 1998) and rat femur (Einhorn et al. 1984, Nottebaert et al. 1989, Ohura et al. 1999) and dog ulna (Key 1934, Heiple et al. 1963, Delloye et al. 1986, 1992, Nilsson et al. 1986, Johnson et al. 1989, Grundel et al. 1991, Schwarz et al. 1991, Cook et al. 1994b). In evaluating the results, various methods of analysis have been used, the principal methods being radiography, histology and torsion testing (Table 2).
Table 2. Summary of the methods of analysis used in segmental bone defect models treated with BMP.
| Implant/carrier | Species | Bone | Defect size | Analysis methods | Authors and year |
|---|---|---|---|---|---|
| bBMP | Rat | femur | 1.0 cm | radiography, histology | Tagaki & Urist 1982 |
| bBMP | Dog | ulna | 2.5 cm | radiography, histomorphometry | Nilsson et al. 1986 |
| bBMP/PLA dBMP/PLA | Dog | radius | 0.3 cm | radiography, histomorphometry | Heckman et al. 1991 |
| rhBMP-2/DBM | Rat | femur | 0.5 cm | radiography, torsion test, histology, radio-isotope boneimaging | Yasko et al. 1992 |
| rhOP-1/collagen | Rabbit | ulna | 1.5 cm | radiography, torsion test, histology | Cook et al. 1994 |
| rhOP-1 /collagen | Dog | ulna | 2.5 cm | radiography, torsion test, histology | Cook et al. 1994 |
| rhOP-1 /collagen | Green monkey | ulna and tibia | 2.0 cm | radiography, torsion test, histology | Cook et al. 1995 |
| rhBMP 2/PGA | Rabbit | ulna | 2.0 cm | radiography, torsion test, histology | Boström et al. 1996 |
| sBMP/TCP | Sheep | tibia | 1.6 cm | radiography, torsion test, histology | Gao et al. 1996 |
| mBMP/coral | Sheep | tibia | 1.6 cm | radiography, torsion test, histology | Gao et al. 1997 |
| bBMP/DBM | Dog | radius | 2.5 cm | radiography, torsion test, histology | Sciadini et al. 1997 |
| bBMP/coral | Dog | radius | 2.5 cm | radiography, torsion test, histology | Sciadini et al. 1997 |
| rhBMP-2 /PLA | Rabbit | radius | 2 cm | radiomorphometry, histomorphometry | Zegzula et al. 1997 |
| rhBMP-2/PLA | Rabbit | radius | 1.0 cm | radiography | Zellin & Linde 1997 |
| rhBMP-2 /PDLLA | Dog | ulna | 2 cm | radiography, histomorphometry | Itoh et al. 1998 |
| rhOP-1 /collagen | Dog | ulna | 2.5 cm | radiography, torsion test, histology | Cook et al. 1998 |
| rhBMP-2 /PDLLA/PGA | Sheep | femur | 2.5 cm | radiography, histology | Kirker-Head et al. 1998 |
| rhBMP-2 /PDLLA/PGA | Rat | femur | 0.5 cm | radiography, torsion test, | Lane et al. 1998 |
| rhBMP-2/PLA/PGA | Rabbit | radius | 2.0 cm | radiography | Texeira & Urist 1998 |
| rhBMP-2/PLA | Rabbit | radius | 2.0 cm | radiomorphometry, torsion test | Wheeler et al. 1998 |
| cBMP/PLA | Dog | radius | 0.3 cm | radiography, histomorphometry | Heckman et al. 1999 |
| rhBMP-2 /PLA/PGA | Rat | femur | 0.5 cm | radiography, histology | Isobe et al. 1999 |
| rhBMP-2/TCP-MCPM | Rat | femur | 0.5 cm | radiography, torsion test | Ohura et al. 1999 |
| rhBMP-2/collagen | Dog | radius | 2.5 cm | radiography, histology, biomechanical testing | Sciadini & Johnson 2000 |
BMP has been used in bone defect models in order to improve bone healing, and the goal has been to achieve equally good or better results with these osteoinductive composite grafts compared to those obtained with the golden standard, autograft bone. Both extracted, naturally occurring BMPs and recombinant BMPs have been used in numerous studies.
Heckman et al. (1991) used both canine and bovine BMP in a relatively small, 3 mm defect in dog radius, using a polylactic acid carrier. The results showed that canine BMP was able to produce a significant increase in new bone formation compared to the controls. In contrast, when bovine BMP was implanted, no significant reparative new bone was found in the defect.
Interestingly, Sciadini et al. (1997a), by using the same model, found out that demineralized bone matrix in combination with bovine BMP resulted in good bone union in all cases, the result being comparable to autograft. In a subsequent study using a natural coral carrier, the results obtained with BMP even exceeded the good results of autografting (Sciadini 1997b).
Stevenson et al. (1994) reported a treatment of a rat femoral defect using bovine BMP with a mixture of hydroxyapatite and tricalcium phosphate used as a carrier. Ceramic together with BMP significantly enhanced the formation of bone in and around the segmental defect.
Gao et al. (1996a) studied sheep BMP with tricalcium and type IV collagen in a sheep tibial segmental defect. After 16 weeks, the BMP group appeared superior in both radiological bone healing and torsional testing of the bone. In a subsequent study, they used moose BMP in the same model and found a larger amount of external callus in the BMP group at 6 weeks. However, after 16 weeks, torsion testing showed lower mechanical strength in the BMP group, and there was also a significantly elevated anti-BMP antibody in serum samples (Gao et al. 1997).
Species-specific canine bone morphogenetic protein induced bone formation in a dog radius bone defect with a PLA/PGA carrier in a study where BMP was also compared with transforming growth factor-beta (TGF-β ). The latter did not induce bone formation in this model (Heckman et al. 1999).
Bovine bone extracted BMP has been widely used in bone defect models because of the good availability of bovine bone. Since bovine BMP cross-reacts immunologically with canine and human BMP and induces heterotopic bone formation in comparable doses in muscle pouches of mice, it is useful in comparative animal research (Nilsson et al. 1986).
Nilsson et al. (1986) used a 2.5 cm segmental canine defect, into which 100 mg of bovine BMP in a capsule was implanted without any carrier material or fixation. On the contralateral side, they implanted a similar capsule with 100 mg of bovine serum albumin. The autograft served as a control in another group of dogs. The results showed that the BMP-treated defects achieved complete regeneration. The incorporation of autogeneic bone in the ulnar defect occurred within the same period as bone regeneration induced by BMP, but the volume of the spindle of callus and the quantity of the BMP-induced new bone were significantly greater than those produced by a cortical bone autograft.
In another study using the same model, however, BMP did not have a positive effect despite the rigid plate fixation (Johnson et al. 1989). Xenogeneic, bovine bone morphogenetic protein (bBMP) and associated insoluble noncollagenous proteins (NCP) were implanted in inbred adult beagle dogs with large, 3–4 cm diaphyseal defects in the ulna with plate fixation. The defects were implanted with either autogeneic cancellous bone grafts (ACG), bBMP/NCP or a composite of ACG and bBMP/NCP with plate fixation. Compared to the restoration of defects implanted with ACG (95%), bone healing occurred in 50 % of the defects implanted with ACG and bBMP/NCP and failed in all defects implanted with bBMP/NCP alone. It was speculated that the interposition of surrounding muscles into the large defect, the species-specific immune response to xenogeneic BMP, the diminished blood supply and the lesser bone regenerative capacity in aged dogs were the reasons for the failure of BMP.
Rat segmental femoral defects 5 mm in length were implanted with two doses of rhBMP-2 (1.4 or 11 g) together with rat DBM carrier (Yasko 1992). Bone formation was demonstrated on the seventh postoperative day in the high-dose rhBMP-2 group, while the low-dose group radiographic showed no evidence of bone in the defects until the third or fourth postoperative week. Radiographic evidence indicated a significant difference between the high-dose and low-dose groups or the bone matrix controls by the ninth week. Mechanically, the healed defects in the high-dose rhBMP-2 group demonstated stiffness comparable to that in the contralateral, intact femora. The established dose-dependent response in the repair of large segmental diaphyseal defects by rhBMP-2 implies that the larger the defect is, the more rhBMP is needed. Isobe et al. (1999) used the same model with a PLA/PGA carrier, and the results showed good bone healing with the rhBMP-2/PLA/PGA capsules compared to the control animals (with a PLA/PGA capsule in the defect), which did not heal.
Using a 1.5 cm segmental rabbit ulnar defect model, the dose-dependent bone-inducing capacity of rhBMP was further confirmed (Cook et al. 1994a). rhBMP-7 at doses from 3.13 to 400 g with allogeneic DBM was implanted in the defects and compared with implants of 250 g of bovine BMP with the same carrier. All of the bovine bone implants and all of the rhBMP-7 implants except those containing 3.13 g of the substance showed complete radiographic osseous union within eight weeks. Histologically, the defect sites were filled with primarily normal lamellar bone with well-developed marrow tissue.
Using a 2 cm segmental rabbit ulnar defect, Boström et al. (1996) were able to demonstrate the same dose-dependent pattern of healing of the defect with rhBMP-2.
Zegzula et al. (1997) treated critical-sized rabbit radial diaphyseal defects with rhBMP-2 in a PDLLA carrier. The BMP they used elicited bone formation and healing of the bone defect.
Recombinant BMP has also been tested in a more demanding, large animal bone defect model (Kirker-Head et al. 1998). A mid-diaphyseal 2.5 cm defect in the sheep femur was stabilized with a plate and implanted with 2–4 mg of rhBMP-2 with a PDLLA carrier. Union occurred in 3/7 of the bones treated with 2 mg of rhBMP-2, 2/3 of the bones treated with 4 mg of rhBMP-2 and none in the control group (no BMP). In the animals that healed, the new bone mineral content equaled that of the intact femur by 16 weeks, and recanalization of the medullary cavity approached completion at 52 weeks. At necropsy, the surgically treated femurs were rigidly healed, the carrier material was resorbed completely, and woven and lamellar bone bridged the defect site. No mechanical testing was performed in this study, as the bones were only tested manually, and the bones with clinical and radiographic union were found to be rigidly healed in the manual test.
Cook et al. (1995) used a bone defect in primates for assessing recombinant BMP-7. They created 2.0 cm defects in the ulnae and in tibiae of African green monkeys and implanted the defects with 250–2000 µg of rhBMP-7 with a bovine collagen carrier. Five of the six ulnae and four of the five tibiae treated with rhBMP-7 exhibited complete healing at six to eight weeks. Histological evaluation revealed the formation of new cortices with areas of woven and lamellar bone and normal-appearing marrow elements. Mechanical testing revealed an average torsional strength to failure of 92 per cent and 69 per cent of that of the contralateral intact ulnae and tibiae, respectively. In this study, rhBMP-7 implants elicited healing of the defects that was as good as or better than that achieved with autogenous bone grafts.
In a recent study, rhBMP-2 combined with bone marrow with a polylactide carrier was implanted in a rat femoral defect (Lane et al. 1999a). The rhBMP-2 and bone marrow composite grafts achieved 100 % union within 6 weeks. The combination was superior compared to each component alone, which strongly supports biologic synergism.
Cook et al. (1994b) used a 2.5 cm segmental canine ulnar defect in a study where recombinant human bone morphogenetic protein (rhOP-1) was used at a dose of 1200 g with a collagen carrier. All defect sites receiving rhOP-1 were completely bridged radiographically by eight weeks. After 12 weeks of implantation, the ulnae had reached mechanical strength comparable to that of a normal ulna.
In a later study by Cook et al. (1998), using the same defect model, ulnae treated with rhOP-1 showed complete radiographic healing at 12 weeks in 89 % of the cases. Histology revealed that the defects were filled with lamellar and woven bone that was in continuity with the host bone, and the mechanical strength of these bones reached 65 % of that of intact ulnae.
Good bone healing was also reported in a study where rhBMP-2 was used with a PLA/PGA/gelatin sponge complex (PGS) as a carrier in canine ulnar defects (Itoh et al. 1998). All defects treated with rhBMP-2 at a dose of more than 160 µg revealed bone union radiographically at 12 weeks, whereas defects treated with PGS alone did not heal.