381

Mechanical testing of recombinant human bone morphogenetic protein-7 regenerated bone in sheep mandibles A Kontaxis1*, M Abu-Serriah2, A F Ayoub2 and J C Barbenel1 1 University of Strathclyde, Glasgow, UK 2 Oral and Maxillofacial Surgery Department, Glasgow Dental Hospital and School, University of Glasgow, UK

Abstract: A new method was developed in this study for testing excised sheep mandibles as a cantilever. The method was used to determine the strength and stiffness of sheep hemi-mandibles including a 35 mm defect bridged by regenerated bone. Recombinant human bone morphogenetic protein-7 (rhBMP-7) in a bovine collagen type-I carrier was used for the bone regeneration. Initial tests on ten intact sheep mandibles confirmed that the strength, stiffness and area beneath the load–deformation curves of the right and left hemi-mandibles were not significantly different, confirming the validity of using the contra-lateral hemi-mandible as a control side. Complete bone regeneration occurred in six hemi-mandibles treated with rhBMP, but the quality and mechanical properties of the bone were very variable. The new bone in three samples contained fibrous tissue and was weaker and less stiff than the contra-lateral side (strength, 10–20 per cent; stiffness, 6–15 per cent). The other half had better-quality bone and was significantly stiffer and stronger (p<0.05), with strength 45–63 per cent and stiffness 35–46 per cent of the contra-lateral side. Hemi-mandibles treated with collagen alone had no regenerated bone bridge suggesting that 35 mm is a critical-size bone defect. Keywords: mandible, cantilever test, bone morphogenetic protein (BMP), bone reconstruction

1 INTRODUCTION The mandible is an important bone in the craniofacial skeleton. It is the most frequently resected bone in cancer surgery and yet it is very difficult to reconstruct due to masticatory stresses, relatively poor blood supply and potential contamination with oral flora [1]. Reconstruction of craniofacial and orthopaedic skeletal defects is a formidable challenge for modern surgery. The objective of this study was to develop a testing method and to investigate the mechanical properties of bone regenerated in defects created in sheep hemimandibles. Jaw function requires adequate strength and stiffness and comparison between the values obtained from the regenerated side and the contra-lateral control side can be used as an indicator of the quality of the regenerated bone. The MS was received on 21 November 2003 and was accepted after revision for publication on 13 July 2004. * Corresponding author: Centre for Rehabilitation and Engineering Studies (CREST), Department of Mechanical and Systems Engineering, University of Newcastle upon Tyne, NE1 7RU, UK. email: [email protected] H09403 © IMechE 2004

1.1 Mechanical test methods The mechanical properties of whole bones can be investigated or prisms of bone can be prepared and tested. The latter method has the advantage that the geometry of the specimens can be regular and controlled by the investigator, but there are, however, significant problems. The bone may undergo changes during preparation, which alter the mechanical properties, particularly the strength. The bone of the mandible is highly inhomogeneous, requiring the preparation of multiple specimens, and the strength and stiffness of the mandible are difficult to determine from the properties of the bone specimens because of the structure of the mandible. It was, therefore, decided to test complete hemi-mandibles. Both three- and four-point bend tests have been used to investigate the mechanical properties of long bones [2–5], but these are difficult to apply in the current study. Beneath the areas of loading and support there are high stresses, which produce significant local deformations and penetration; these stresses would produce even greater effects in the softer regenerating bone [6 ]. The space available in the mandible was also limited to a defect of 35 mm. The mandibles were therefore tested as Proc. Instn Mech. Engrs Vol. 218 Part H: J. Engineering in Medicine

382

A KONTAXIS, M ABU-SERRIAH, A F AYOUB AND J C BARBENEL

cantilevers, with the direction of loading normal to the teeth, mimicking the loading that occurs during masticatory function. 1.2 Bone regeneration Bone defects caused by trauma or surgery can be repaired by new bone regenerating from intact bone and the periosteum if the defect is less than a critical size. Defects greater than the critical size generally require bone grafting, but surgical reconstruction of defects, particularly defects created by cancer surgery, is often compromised by the limitation on the availability of donor tissues and wound healing problems [7]. There is an increasing understanding of the biochemical factors that control and modulate bone induction and growth and the use of these factors, particularly bone morphogenetic proteins (BMPs), to stimulate bone formation is a realistic alternative to the use of autogenous bone grafts [8]. To date, at least six human BMPs have demonstrated osteogenic activity: BMP-2, BMP-4, BMP-5, BMP-6, BMP-7 (also referred to as osteogenic protein (OP)-1) and BMP-8 (OP-2), some of which are available in commercial quantities through genetic engineering. All BMPs require a carrier to act as a delivery system. Although the use of BMPs is promising, it is at an early stage and occasionally associated with unpredictable outcome and prohibitive cost that limit their widespread clinical use [9]. 2 MATERIALS AND METHODS The method developed in this study evaluates the biomechanical properties of the operated side (OS) hemimandibles by direct comparison with the contra-lateral, non-operated side (NOS). To validate the method, the properties of both sides of ten NOS sheep mandibles were compared in order to confirm that the contralateral hemi-mandibles were sufficiently alike for the intact side to act as a control for the operated side. Despite the great differences in mastication pattern between humans and sheep [10] the latter is a large mammal and has a similar bone formation rate to a human [11]. The species has been used extensively in similar bone research [12, 13] and is known to have a lower jaw the size of which closely resembles that of humans [14]. In this study, a full thickness continuity defect 35 mm long was prepared in the toothless region (diastema) between the right fourth incisor and the first premolar of the mandible of 12 adult sheep (Fig. 1). This also involved the excision of the corresponding part of the inferior dental and mental neurovascular bundle. Few studies have explored critical size mandibular defects [15] and the size of the bone defect in this study was believed to be non-healing. Proc. Instn Mech. Engrs Vol. 218 Part H: J. Engineering in Medicine

Fig. 1 A schematic side view of the front of the sheep’s mandible showing the surgical model used (A, front; P, back; ID, inferior dental )

The animals were divided into two groups: (a) a control (Ctrl ) group of six animals in which only collagen was implanted and (b) a BMP- group of six animals in which 1 mg/cm3 of recombinant human bone morphogenetic protein-7 (rhBMP-7) in bovine collagen type-I carrier was applied; a dose of between 0.6 and 2.5 mg/cm3 has been reported to produce the best bone formation [2, 8, 16 ]. The Ctrl group was included to identify the boneforming capacity of bovine collagen type I and to prove that the size of the defect is critical (i.e. lacks the capacity to heal spontaneously). A custom-made metallic bone plate was used to bridge the surgical defect and to maintain the continuity of the mandibular bone structure, allowing normal feeding soon after the surgery (Fig. 2a). The plate was fixed in place using six screws (Fig. 2b). The animals were sacrificed 3 months after the surgery and the mandibles were then explanted. The metal plate was carefully removed and mandibles were split into OS and NOS hemimandibles. The biomechanical properties of the OS were compared with those of the contra-lateral NOS, using a cantilever test. 2.1 Mechanical test mode The bone defect had been created in the diastema between the fourth incisor and the first premolar (Fig. 1). It was, therefore, necessary to determine the mechanical properties of the bone in this region. The mandibular teeth of the sheep are supported by the bone of the horizontal ramus (known also as the body of the mandible). The shape and size of the teeth, particularly the first premolar, are variable and all the posterior teeth have large roots. As a result the amount of bone around these roots is less than in other areas of the horizontal ramus. Pilot tests showed that gripping the horizontal ramus distal to the first premolar and applying loading anteriorly produced failure through the bone supporting the molar roots, rather than the bone at the diastema. It was, therefore, necessary to develop a device to support the mandible in the premolar and molar region. H09403 © IMechE 2004

MECHANICAL TESTING OF rhBMP-7 REGENERATED BONE

383

Fig. 3 Testing a specimen. The hemi-mandible was clamped in the two-part jig with the lingual side of the specimen vertical and loaded anterior to the toothless diastema. A is the rectangular frame holding the bone at the area of the premolar and molar teeth, B supports the mandible distally and prevents rotation and C is the pushing rod connected to the load cell

Fig. 2 (a) A photograph showing the defect and the metallic plate used over the defect to hold the bone segments. (b) A radiograph of the defect and the plate fixed in place with screws

2.2 Mechanical testing The hemi-mandible under test was held in a jig that was specially designed using AutoCADA. A three-dimensional model of a sheep hemi-mandible was created for optimizing the jig design. The main aim of the designing was to hold and stabilize the hemi-mandible, both horizontally and vertically, around the area of the first premolar and distally. The mandible was supported by two components fixed to an adjustable base (Fig. 3). A rectangular aluminium frame (Fig. 4) was designed to hold the bone in the area of the premolar and molar teeth. The lower border of the mandible was located on the base of the frame and the mandible and teeth were stabilized vertically in the frame by two screwed rods (Fig. 4b). The upper surfaces of the second premolar and molars were at the same level and these teeth were loaded by a rod that terminated in a plate that was free to rotate both about and normal to the rod (Fig. 4a). The first premolar had variable position and size and was generally located below the level of the more distal teeth; a second rod, terminating in an inverted cone was, therefore, used to stabilize this tooth. The lateral position was maintained by rods, terminating in a swivelling plate, on each side of the mandible. A flat metal plate was interposed between the swivelling H09403 © IMechE 2004

Fig. 4 The vertical location of the premolar and molar teeth is maintained by two threaded rods (A) and the lateral location by a threaded rod on each side (B)

plate and the inner and outer surfaces of the mandible. Close adaptation between the flat metal plate and the underlying bone surface was achieved using a ‘chemical metal’ (Loctite Technology ProductsA). Care was taken to ensure that lateral support had a clearly defined anterior limit, matching the aluminium frame (Fig. 4). The second component of the jig supported the mandible Proc. Instn Mech. Engrs Vol. 218 Part H: J. Engineering in Medicine

384

A KONTAXIS, M ABU-SERRIAH, A F AYOUB AND J C BARBENEL

distal to the horizontal ramus, preventing bending moments from rotating the specimen (Fig. 3b). Initial tests showed that the mandibular incisors were displaced under load and their supporting tissues failed before the mandible. Vertical loads were, therefore, directly applied to the mandibular bone just in front of the anterior margin of the surgical defect by a bar that was shaped to reduce penetration into the bone. The distance between the point of loading and the fixed point at the front edge of the rectangular frame at the first premolar was standardized to 35 mm. The jig holding the mandible was mounted on the moveable cross-head of an InstronA 4505 series test machine (Instron, High Wycombe, UK ). A rod carrying the loading bar was connected to a compressive load cell. An initial preload of 15 N was applied, after which the cross-head was driven at 5 mm/min towards the loading bar. The load and cross-head displacement were being continually recorded. 2.3 Test specimens All specimens were supplied as hemi-mandibles, which were stored at temperature of −24 °C. The specimens were defrosted before testing, stripped of soft tissue and mounted for testing. 3 RESULTS 3.1 Intact mandibles Cantilever bending tests were made on the left and right hemi-mandibles of ten normal mandibles. The typical load–displacement response (Fig. 5) had an initial nonlinear region at the initial stages of the displacement. A linear region was followed by a decrease in the gradient, which terminated by an abrupt fall in load as the mandible fractured. In all tests, failure took place anterior to the first premolar. In some tests, failure occurred at the maximum load (Fig. 5, left), but in several specimens a small load decrease occurred prior to failure (Fig. 5, right). The gradient of the linear part of the load–displacement curve was calculated as it indicates the stiffness of the mandible. The maximum applied load was identified and translated into maximum applied moment, as a

Fig. 5 Typical load–displacement curves obtained by testing the right and left sides of an intact mandible. The vertical bars represent the resolution of the load cell

measure of the strength of the mandible. The area beneath the load–displacement curve was calculated since it expresses the energy absorbed by the bone during the bending test. The mean values of the three variables (maximum applied moment, elastic stiffness and energy absorbed) and the ratio of these values for the right versus the left hemi-mandible are shown in Table 1. The crack that initiated failure in the specimens appeared at the lower border of the mandible, very close to the site of the maximum applied moment, which was 35 mm away from the loading point. Fracture lines occurred transverse to the body of the mandible and these had a ragged fracture surface. Fracture lines at 45° to the mandibular body with a smooth fracture surface were also observed. In all the specimens the crack propagation could be heard. 3.2 Operated mandibles The right hemi-mandibles (OS), in which the defect had been filled with collagen and BMP, had bone bridging the defect, although in some specimens the shape of the regenerated bone was different from the normal mandible. The load–displacement behaviour on the left (NOS) was similar to that of the intact hemi-mandibles. The

Table 1 Mean values of parameters and ratios from tests on ten pairs of intact hemimandibles Mean values (Standard error) Parameter

Right side n=10

Left side n=10

Mean values of right-to-left ratio (Standard error) n=10

Maximum applied moment Stiffness Energy absorbed

50.31 (3.11) N m 0.33 (0.02) N/m 5.34 (0.37) J

50.62 (3.04) N m 0.37 (0.03) N/m 5.02 (0.54) J

0.996 (0.023) 0.942 (0.083) 1.136 (0.091)

Proc. Instn Mech. Engrs Vol. 218 Part H: J. Engineering in Medicine

H09403 © IMechE 2004

MECHANICAL TESTING OF rhBMP-7 REGENERATED BONE

data followed a normal distribution ( Kolmogorov– Smirnov normality test; p=0.15) and a t test suggested that the NOSs were slightly stiffer and stronger [rejecting equal means at 99 per cent confidence level; p=0.033 (Table 2)]. The crack initiation and propagation also followed the pattern of the intact hemi-mandibles. The OSs were all weaker and less stiff than the contralateral NOSs (p<0.0005) and there was considerable variability. While some specimens failed at displacements similar to normal values with a loading pattern similar to NOSs, others were considerably weaker and failed after much greater displacement under a very low load (see Fig. 7 later). Unlike the NOSs and the intact hemi-mandibles, in the rhBMP-7-treated OS hemi-mandibles, crack initiation occurred at the upper surface of the regenerated bone close to the interface with the normal bone and propagated vertically. Five out of six NOS hemi-mandibles of the specimens in which the defect had been filled with collagen alone, once again, showed close resemblance to the intact hemimandibles (p=0.719). The data on the sixth specimen were not valid due to testing problems and they were not taken into account. The OS side of the Ctrl group failed to achieve complete bone healing and the hemimandible deformed greatly on the application of even low loads; it was not possible to calculate the stiffness and strength of these specimens. The failure to achieve mechanically detectable bone regeneration clearly indicated that the 35 mm bone gap was a critical size defect in which complete bone repair was not possible. Results of the tests on the operated mandibles showed a strong linear relationship between strength and stiffness for both the non-operated and regenerated bone (Fig. 6), with a correlation coefficient of 0.932. Such correlation was much weaker for all the intact hemimandibles, with the coefficient being 0.433.

4 DISCUSSION The mechanical test method was simple and reliable and was able to detect and assess quantitatively the different quality of the bone reconstructed using BMP. The canti-

385

Fig. 6 The linear correlation for the strength–stiffness plot was very high for both the OSs and the NOSs of the BMP group mandibles

lever test resembled the loading of the sheep mandible during chewing. The mandibular joint localizes the posterior part of the bone, the masseter and other muscles of mastication generate upward forces, and downward forces act on the incisor teeth. Neither of the mean right-to-left ratios for the ten normal mandibles in Table 1 was significantly different from 1 at 95 per cent confidence level (p=0.810). These results confirm that values of mechanical parameters obtained from the NOS can be used as a control to compare with values obtained from the contra-lateral OS. The parameter values of the bone on the NOS of the BMP and collagen groups of test animals were not significantly different (p=0.085) and the results were pooled (Table 2). The pooled values of the stiffness, strength and energy parameter in the NOS of the test animals were not significantly different from the values in Table 1 for the intact mandibles but paradoxically, when the values for the NOSs of the animals implanted with BMP were taken, they alone were significantly different (p=0.004) from those in Table 1. These differences may be the result of greater than normal chewing on the NOS side, producing adaptive remodelling of the bone [17–20]. The biomechanical parameters of the regenerated bone can be compared with either the

Table 2 Mean values of parameters and ratios from operated mandibles. Results for 11 NOS hemi-mandibles and six mandibles treated with BMP and collagen were compared Mean values (Standard error) Parameter

Right OS n=6

Left NOS n=11*

Mean values of right-to-left ratio (Standard error) n=6

Maximum applied moment Stiffness Energy absorbed

22.24 (4.91) N m 0.15 (0.04) N/m 2.95 (0.29) J

59.12 (3.54) N m 0.61 (0.04) N/m 3.05 (0.82) J

0.363 (0.081) 0.244 (0.070) 0.606 (0.073)

* Six NOSs of the mandibles treated with rhBMP-7 and collagen+five NOSs of the mandibles treated with only collagen H09403 © IMechE 2004

Proc. Instn Mech. Engrs Vol. 218 Part H: J. Engineering in Medicine

386

A KONTAXIS, M ABU-SERRIAH, A F AYOUB AND J C BARBENEL

parameters of the contra-lateral NOS or of the intact mandibles, without changing the general qualitative conclusions. It is, however, important to use the appropriate contra-lateral side to compare the quantitative mechanical properties of the regenerated bone. The strengths and stiffnesses of the BMP regenerated bone (Table 2, second column) were lower and more variable than the corresponding values for both the NOSs of the treated and of the ten intact mandibles. The appearance of the BMP regenerated bone was also variable, but the specimens appeared to fall into two groups. Three of the specimens showed regenerated bone with relatively high stiffnesses and strengths. The stiffness ratios were 0.347–0.467, compared with the pooled value of 0.244 in Table 2; strength ratios were 0.466–0.633 (pooled value of 0.363). Failures occurred by fast crack propagation that was accompanied by an abrupt fall in load (Fig. 7a). The cracks propagated vertically to the body of the mandible and the noise of crack propagation could be heard, as in normal bone.

Fig. 7 There were two qualities of bone regeneration of the BMP group: (a) some of the hemi-mandibles failed at displacements similar to normal values; (b) others, containing soft tissue, were considerably weaker and failed after much greater displacement (b). The right shows the OS and the left the NOS Proc. Instn Mech. Engrs Vol. 218 Part H: J. Engineering in Medicine

The regenerated tissue in the remaining three specimens was a mixture of bone and soft fibres that were visible especially in the area underneath the site occupied by the metallic plates. The stiffness ratios were only 0.055–0.175 and there was considerable extension before failure (Fig. 7b). The regenerated tissue was also weak, with strength ratios of 0.092–0.249. During failure, cracks slowly propagated vertically, while the force remained almost constant. No noise of crack propagation could be heard. The energy-absorbed ratio failed to distinguish between the different quality of the regenerated bone, because of the different shapes of the load–displacement curves. Although the weaker specimens failed under much lower force, this was accompanied by greater deformations, giving areas beneath the curve that was, in some cases, similar to those of the stronger specimens. The variable shape of the individual mandibles was accommodated by the multiple adjustments of the holding device. The variability of the shape and the thickness of the cortical plate made loading the upper surface of the mandible difficult. The loading bar compressed any residual soft tissue before making full contact with the upper surface of the hemi-mandible and generally produced a local deformation of the bone surface. These affects produced an initial characteristic non-linearity in the load–displacement curve in most tests. However, the non-linear section was omitted when calculating the slope to obtain the stiffness of the specimen. The cross-section of the NOS hemi-mandible showed that the shape and bone distribution were non-uniform and the lower part of the mandible contained the inferior (alveolar) canal, suggesting that the neutral axis would be located in the upper half of the cross-section of the mandible. The compressive stress at the lower border of the mandible would, therefore, be greater than the stress at the upper border, which may account for the site of crack initiation, which was located in the lower border of the mandible and very close to the site of the maximum applied moment. All the cracks propagated rapidly, but the direction of crack propagation was variable, suggesting propagation under either shear or normal stresses. In bone, the resistance to crack growth under shear loading and its relation to crack growth under tensile loading is not well understood [21]. Unlike the NOS the initiation of the crack was in the upper surface of the regenerated bone, close to the interface with the normal bone. The regenerated bone in the lower part of the mandible and the absence of the inferior canal in these specimens suggest that the neutral axis would be located below the horizontal midline of the transverse section of the specimen, producing the largest stress at the upper border of the mandible. There were also histological differences between the two qualities of regenerated bone. The bone of the specimens with better mechanical properties was more compact and less porous, and the inner and outer cortices H09403 © IMechE 2004

MECHANICAL TESTING OF rhBMP-7 REGENERATED BONE

were similar to the normal mandible. The poorer-quality group contained less mature bone with more narrow spaces (Fig. 8). The applications of BMPs are popular not only in dentistry and mandibles but also in treatment of fractures cartilage defects and even arthritis [22]. Most of the studies, however, relate to small bone defects that would heal without the use of BMPs. Some recent animal studies have demonstrated the potential of BMPs to enhance spinal fusion and to repair critical-size defects in long bones. There are only a few reports describing reconstruction of critical-size defects in mandibles and most of these confirm the variability of the quality of the regenerated bone shown in the current study [23, 24].

5 CONCLUSIONS The method of cantilever bending allowed the mechanical properties of the body of sheep mandibles to be evaluated. The use of screwed rods enabled hemimandibles and molar teeth of different shape and size to be successfully supported for testing. Loading the incisor teeth was unsatisfactory because the supporting tissues

387

of the teeth failed at lower loads than the mandible. It was necessary to load the bone of the mandible distal to the incisor teeth, but compression of residual soft tissue and local bone distortion produced artefactual initial results. The strength, stiffness and energy values obtained from the ten intact mandibles and 11 NOS hemimandibles all showed good reproducibility, although the coefficient of variability for the energy absorbed was larger than that of the other parameters. The mean values of these three parameters for the left and right intact hemi-mandibles were very similar and none of the right-to-left ratios was significantly different from 1. The mean stiffness and strength of the bone in the NOS of the test animals were significantly greater than the values obtained from the intact mandibles, showing the importance of using appropriate contra-lateral controls to assess the mechanical properties of the regenerated bone. In the OS hemi-mandibles, regenerated bone bridged the surgical defect where it had been filled with rhBMP-7 and collagen, but there was no bone union in those in which only collagen had been used, confirming that 35 mm is a critical-size defect and that collagen alone does not have a strong bone-forming capacity. The shape of the BMP regenerated mandibular segments and the appearance of the regenerated bone were variable. The regenerated bone present in three of the specimens had a stiffness and strength that were a significant proportion of the values obtained from the contra-lateral NOS. The regenerated tissue in three specimens was a mixture of bone and fibrous tissue and had very poor mechanical properties. The cross-sectional shapes of the intact and regenerated mandibles, and hence the position of the neutral axes, were different. In the intact mandibles, failure was initiated at the lower border of the mandible, but failure was initiated at the upper border in regenerated bone. ACKNOWLEDGEMENTS The experimental work was carried out in the Bioengineering Unit, University of Strathclyde, and the authors thank David Smith for his technical assistance. This work was supported by a research grant from Stryker–Leibinger. REFERENCES

Fig. 8 Two different qualities of regenerated bone were observed: (a) better-quality bone was less porous and showed attempts at restoring the inner and outer cortices (arrows); (b) poor-quality bone was bulbous, containing more marrow spaces and immature bone formation H09403 © IMechE 2004

1 Hollinger, J. O. and Kleinschinidt, J. C. The critical size defect as an experimental model to test bone repair materials. Craniofacial Surg., 1990, 1, 60–68. 2 Cook, S. D., Wolfe, M. W., Salkeld, S. L. and Rueger, D. C. Effect of recombinant human osteogenic protein-1 on healing of segmental defects in non-human primates. J. Bone Jt Surg., 1995, 77A, 734–750. 3 Ruff, C. B. and Hayes, W. C. Cross-sectional geometry of Proc. Instn Mech. Engrs Vol. 218 Part H: J. Engineering in Medicine

388

4

5

6

7 8

9

10 11

12

13

A KONTAXIS, M ABU-SERRIAH, A F AYOUB AND J C BARBENEL

Pecos Pueblo femora and tibiae – a biomechanical investigation. I: method and general patterns of variation. J. Phys. Anthropology, 1983, 60, 359–381. Raftopoulos, D. D. and Qassem, W. Three dimensional curved beam stress analyses of the human femur. Am. J. Phys. Anthropology, 1983, 60, 383–400. Bertram, J. E. and Biewener, A. A. Bone curvature: sacrificing strength for load predictability. J. Theor. Biol., 1988, 131, 75–92. Lovejoy, C., Burstein, A. and Heiple, K. The biomechanical analysis of bone strength: a method and its application to platycnemia. Am. J. Phys. Anthropology, 1976, 44, 489–505. Rudkin, G. H. and Miller, T. A. Growth factors in surgery. Plast. Reconstr. Surg., 1996, 97, 469–476. Ripamonti, U., van den Heever, B., Sampath, T. K., Tucker, M. M., Rueger, D. C. and Reddi, A. H. Complete regeneration of bone in the baboon by recombinant human osteogenic protein-I (hOP-1, bone morphogenetic protein-7). Growth Factors, 1996, 123, 273–289. Muthukumaran, N., Ma, S. and Reddi, A. H. Dosedependence of and threshold for optimal bone induction by collagenous bone matrix and osteogenin-enriched fraction. Collagen Related Res., 1988, 8, 433–441. Nanamaker, D. M. Experimental models of fracture repair. Clin. Orthop. Related Res., 1998, 355S, S56–S565. den Boer, F. C., Patka, P., Bakker, F. C., Wippermann, B. W., van Lingen, A., Vink, G. Q., Boshuizen, K. and Haarman, H. J. New segmental long bone defect model in sheep: quantitative analysis of healing with dual energy x-ray absorptiometry. J. Orthop. Res., 1999, 17, 654–660. Boyne, P. J. Animal studies of applications of rhBMP-2 in maxillofacial reconstruction. Bone (Suppl.), 1996, 83S–92S. Hanisch, O., Tatakis, D. N., Rohrer, M. D., Wohrle, P. S., Wozney, J. M. and Wikesjo, I. M. Bone formation and osseointegration stimulated by rhBMP-2 following subantral augmentation procedures in non human primates. J. Oral Maxillofacial Implants, 1997, 12, 785–792.

Proc. Instn Mech. Engrs Vol. 218 Part H: J. Engineering in Medicine

14 Wittenberg, J. M., Mukherjee, D. P., Smith, B. R. and Kruse, R. N. Biomechanical evaluation of new fixation devices for mandibular angle fractures. Int. J. Oral Maxillofacial Surg., 1997, 26, 68–73. 15 Salmond, R. and Duncan, W. J. Determination of the critical size for non-healing defects in the mandibular bone of sheep. Part I: a pilot study. J. N. Z. Soc. Periodontology, 1997, 81, 6–15. 16 Cook, S. D., Salkeld, Sl. and Rueger, D. C. Evaluation of recombinant osteogenic protein-1 (rhOP-1) placed with dental implants in fresh extraction sites. J. Oral. Implantology, 1995, 21, 281–289. 17 Sugiyama, T., Taguchi, T. and Kawai, S. Adaptation of bone to mechanical loads. Lancet, 2002, 30, 359(9312), 1160. 18 Akhter, M. P., Cullen, D. M. and Recker, R. R. Bone adaptation response to sham and bending stimuli in mice. J. Clin. Densitometry, 2002, 5(2), 207–216. 19 Cowin, S. C. Bone stress adaptation models. J. Biomech. Engng, 1993, 115(4B), 528–533. 20 Cowin, S. C. Mechanical modeling of the stress adaptation process in bone. Calcification Tissue Int., Suppl. 1, 1984, 36, S98–S103 21 Norman, T. L., Nivargikar, S. V. and Burr, D. B. Resistance to crack growth in human cortical bone is greater in shear than in tension. J. Biomechanics, 1996, 29, 1023–1031. 22 Issack, P. S. and DiCesare, P. E. Recent advances toward the clinical application of bone morphogenetic proteins in bone and cartilage repair. Am. J. Orthop., 2003, 32(9), 429–436. 23 Toriumi, D., Kotler, H., Luxenberg, D., Holtrop, M., and Wang, E. Mandibular reconstruction with a recombinant bone-inducing factor. Archs Otolar. Head Neck Surg., 1991, 117, 1101–1112. 24 Saadeh, P. B., Khosla, R. K., Mehrara, B. J., Steinbrech, D. S., McCormick, S. A., DeVore, D. P. and Longaker, M. T. Repair of a critical size defect in the rat mandible using allogenic type I collagen. J. Craniofacial Surg., 2001, 12(6), 573–579.

H09403 © IMechE 2004