- Email: [email protected]
Experimental model of tibial plateau fracture for biomechanical testing
ARTICLE IN PRESS
Journal of Biomechanics 39 (2006) 1355–1360
Short communication
www.elsevier.com/locate/jbiomech www.JBiomech.com
Experimental model of tibial plateau fracture for biomechanical testing Ahmad M. Alia, Michael Salehb, Stefano Bolongarob, Lang Yangb, a
Department of Orthopedics and traumatology, Oldchurch Hospital, Waterloo Road, Romford, Essex RM7 0BE, UK Orthopaedic and Traumatic Surgery Research Group, Division of Clinical Sciences (North), University of Sheffield, Northern General Hospital, Sheffield S5 7AU, UK
b
Accepted 11 March 2005
Abstract Although adequate reduction and stable fixation have been recognized to be the prime goals in the treatment of displaced tibial plateau fractures, the optimal fixation technique remains controversial. The lack of a reliable model and a standard methodology contribute to this situation. The purpose of this study is to develop an experimental model of a tibial plateau fracture and a testing methodology that reproduces the failure mode commonly seen in the clinical setting. Using solid-foam and composite Sawbones tibiae, three different models of bi-condylar tibial plateau fracture (solid-foam, reinforced solid-foam and composite), six specimens for each model, were created and stabilized with double plating. The specimens were subjected to cyclic axial compression with increasing maximum load until failure. A femoral component of a total knee replacement of similar size and shape to the synthetic tibial surface was used as a load applicator. The experiment was repeated on six specimens of human cadaver tibiae. Among the Sawbones specimens, only the reinforced solid-foam model was found to produce a consistent failure mode (collapse in the medial plateau) comparable to that reported clinically in the literature. This mode of failure was also confirmed by the cadaver experiments. The failure load of the reinforced solid-foam model ranged from 4150 to 4260 N with a mean7SD of 4201744 N and a coefficient of variance of 0.01, whereas for the cadaver model the failure load ranged from 1675 to 6096 N with a mean7SD of 376871482 N and a coefficient of variance of 0.39. We recommend the reinforced-foam model for future mechanical tests to compare different fixation methods for tibial plateau fractures. r 2005 Elsevier Ltd. All rights reserved. Keywords: Biomechanical testing; Experimental model; Tibial plateau fracture; Sawbones
1. Introduction Tibial plateau fracture is a common injury, with an incidence of 9.2% of all tibial fractures (Burri et al., 1979; Fernandez, 1988; Hohl, 1967; Roberts, 1968). Although adequate reduction and stable fixation have been recognized to be the primary goals in the treatment of displaced tibial plateau fractures, the optimal fixation technique remains controversial (Blokker et al., 1984; Hohl, 1967; Schatzker et al., 1979; Waddell et al., 1981). Corresponding author. Tel: +44 114 2714900; fax: +44 114 2619246. E-mail address: l.yang@sheffield.ac.uk (L. Yang).
0021-9290/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbiomech.2005.03.022
Loss of reduction is frequent and may depend on the type of fixation (Benirschke et al., 1992; Bowes and Hohl, 1982; Delamarter et al., 1990; Hohl, 1967; Marsh et al., 1995). The biomechanical data based on cadaveric bone have failed to show conclusively which fixation method is best for each type of tibial plateau fracture (Denny et al., 1984; Horwitz et al., 1999; Koval et al., 1996; Parker et al., 1999). Furthermore, conflicting conclusions were reached in two different studies which tested the same type of fracture using the same fixation technique (Denny et al., 1984; Koval et al., 1996). The use of cadavers for mechanical testing has merits and drawbacks. The obvious advantages are the anatomical and mechanical reality. The drawbacks are
ARTICLE IN PRESS 1356
A.M. Ali et al. / Journal of Biomechanics 39 (2006) 1355–1360
mainly the difficulties in obtaining a sufficient number of specimens (Earnshaw, 2000) (Einhorn, 1992; Goldstein, 1987) and the wide variability of bone quality (mechanical property) (Einhorn, 1992; Goldstein, 1987). The wide variability of bone quality leads to a wide variability in mechanical quantity testing, which in turn demands a greater number of specimen to ascertain a satisfactory statistical power. Without quantifying bone quality and random assignment, it is possible that one fixation method is tested mainly on specimens of poor bone quality and another on good quality bones. This may explain the differences found in the studies by Denny and Koval (Denny et al., 1984; Koval et al., 1996). Bone mineral density can be measured readily by dual energy x-ray absorptiometry (DXA), but the density on its own is not totally representative of bone quality (Sievanen et al., 1994). Other variables, like orientation, thickness and separation of the trabeculae are also important factors that determine the mechanical properties of cancellous bone (Einhorn, 1992; Goldstein, 1987). Unfortunately, these parameters are much more difficult to quantify. With a limited number of specimens and without proper control of bone quality between the comparative groups, the validity of the conclusions from previous studies is questionable (Earnshaw, 2000). The method used in biomechanical testing is another important variable that makes it difficult to compare the results of different studies. Parker et al. (1999) used a polyurethane load applicator to ramp compress the lateral plateau, and the yield force was defined as the force at 2 mm plateau subsidence. Horwitz et al. (1999) and Koval et al. (1996) used a steel ball as load applicator, ramp and cyclically loaded the specimen, and the total subsidence between fixation methods were compared. There is a need for a reliable model of tibial plateau fracture for mechanical testing, as well as a standard methodology permitting comparison to be made between different studies and of different fixation techniques. Synthetic bone material has been used previously in the mechanical testing of fixation devices (Gerber and Ganz, 1998; Prayson et al., 2001). Sawboness limb models (Sawbones Europe AB, Sweden) have been shown to have similar material and structural properties to that of human bone (Cristofolini and Viceconti, 2000; Sawbones Europe AB., 2002). The obvious advantages of using Sawboness are the assurance of uniformity, consistency in its material and geometry, and therefore its mechanical properties. This synthetic bone model has to be compared to a cadaveric model to validate the mode of failure. With such a model, the performance of fixation devices can be evaluated more reliably since the variability within specimens is properly controlled. As the first phase of an investigation into the effects of bone quality on the strength of different fixation
methods for tibial plateau fracture, the aim of this study is to develop a model of tibial plateau fracture and a testing methodology that consistently produces the failure mode commonly seen in clinical settings.
2. Materials and methods Using Sawboness tibia three different models were employed in the first part of this study (Fig. 1): Model 1—solid-foam bone (Sawboness Product no. 1116): a rigid polyurethane foam material throughout that simulates cancellous bone. It has a circular ‘‘medullary canal’’ of 12 mm in diameter. Model 2—composite bone (Sawboness Product no. 3301): the core of the bone is made of rigid polyurethane foam and the cortex layer is formed by pressure injecting a mixture of short e-glass fibres and epoxy resin around the form core. The structural properties of the whole tibial model have been validated (Cristofolini and Viceconti, 2000). Model 3—reinforced solid-foam bone: the same as the first model but reinforced with an ‘‘intramedullary’’ carbon-fibre rod (Tufnols 10G/40, Tuflon Ltd, Birmingham, UK) to support the diaphysis. The rod was 12.7 mm (0.5 in) in diameter and inserted into the tibial canal about 85 mm distal to the plateau and just distal to the most distal screws of the fixation plate. This model was created specifically for this study. The mechanical properties of the Sawbones materials were found to be in the range of natural bones (Currey, 1984; Sawbones Europe AB., 2002). The Sawbones were obtained from one manufacturing batch ensuring the same material properties and geometry. Six specimens of each model were used. A bi-condylar tibial plateau fracture (AO/OTA type 41-C1, Schatzker type-V (Schatzker et al., 1979)) was created using an automated thin blade saw (Fig. 1A). A template was used to create a reproducible cut with the same inclination and distances from the tibial spine to the lateral and medial edges of the tibial plateau. The fractures were reduced and fixed using the AO dual plating technique, this was performed by a single orthopaedic surgeon (AMA) who was fully trained in the AO principles of internal fixation (Schatzker, 1987) (Fig. 2). This dual plating technique involved two L shaped buttress plates (six hole), one for each condylar fracture, with three partially threaded 6.5 mm cancellous screws proximally and three 4.5 mm cortical screws distally in each plate. The model was fixed in an engineering angle vice, via a split resin mould that held the distal tibia rigidly, with the long axis of the tibial shaft at 31 varus and a 101 posterior slope to simulate the normal alignment in midstance full weight bearing (Fig. 2). The specimen was then mounted and fixed onto a material-testing machine (Autograph ASG10kN, Shimadzu Corporation, Kyoto,
ARTICLE IN PRESS A.M. Ali et al. / Journal of Biomechanics 39 (2006) 1355–1360
1357
Fig. 1. (A) Bi-condylar tibial plateau fracture (Schatzker Type-V). (B) Model 1: solid-foam bone. (C) Model 2: composite bone. (D) Model 3: reinforced solid-foam bone with an intramedullary carbon-fibre rod (pointed by an arrow) to support the diaphysis.
Fig. 2. (A) Synthetic and (B) cadaver bi-condylar tibial plateau fracture models instrumented and ready for mechanical testing. The arrows point to the sites of the stress riser where the failure occurred in Model 1.
Japan). In order to measure the collapse of the two plateaus under loading, four extensometers, two for each plateau, were mounted on the model anteromedially, postero-medially, antero-laterally and pos-
tero-laterally (Fig. 2). These extensometers were calibrated against the crosshead movement (0.001 mm resolution) and found to be accurate within 5%. In the process of developing a suitable loading applicator, different applicators described in other studies were tried (Horwitz et al., 1999; Koval et al., 1996; Parker et al., 1999). However, these applicators were discarded as they were found to cause a focal crushing effect localized to one area, instead of collapse in the plateau fixation. Furthermore, some of these applicators (spherical ball) are designed to test only one plateau. Therefore, a femoral component of a total knee replacement of a similar size and shape to the tibial surface was finally selected and used (Size 4 PFCs SigmaTM Keen System, DePuy Orthopaedics Inc., Warsaw, USA) as a loading applicator in order to deliver forces on both tibial plateau surfaces simultaneously (Fig. 2). Initial testing revealed that as loading increased to 2000 N, a consistent posterior slide or translation of the femur on the tibial articular surface occurred. This was due to the posterior tibial slope and the low friction between the two surfaces. To avoid this sliding, we inserted two small sharp spikes (2 mm) on the articular surface of the femoral prosthesis, one in each condyle to engage the tibial surface during testing (Fig. 2). This proved to be effective in preventing any sliding or subluxation during axial loading. Cyclical compression was applied to the specimen at a rate of 20 mm/min. For conditioning the system four
ARTICLE IN PRESS 1358
A.M. Ali et al. / Journal of Biomechanics 39 (2006) 1355–1360
loading cycles at 500 N were first performed. The actual tests started with four loading cycles at 1000 N, followed by four loading cycles at 2000 N, four cycles at 4000 N and four cycles at 9000 N, if there was no obvious failure of the construct during the previous loading level, otherwise the experiment stopped at the load level where an obvious failure of the construct occurred. The output from the material testing machine (load and crosshead displacement) and extensometers were recorded for later analysis at 10 Hz using a computerized data acquisition system. Failure was defined as a plastic displacement of 3 mm or more in the articular surface in either condyle. A previous study has shown that 3 mm represented a significant step-off displacement, which is unacceptable clinically (Ali et al., 2002). In the second part of the study, six human cadaveric tibiae of similar size and bone density were used. The tibiae were collected from subjects without a medical history of skeletal pathology (mean7SD age ¼ 66.57 10 years, range 53–78 years). Bi-planar radiographs were used to exclude those specimens with other pathological lesions. Following soft tissue removal, they were wrapped in saline soaked gauze sponges and stored in plastic bags at 70 1C. They were thawed in room temperature before experiment. DXA scanning using Hologic QDR 4500 was performed on all specimens, and it confirmed that these tibiae are a homogenous group with an average BMD of 0.65 g/cm2 (range 0.55–0.75 g/cm2; SD 0.08 g/cm2). The method of creating and stabilizing the fracture and the protocol of mechanical testing was identical to that used for the synthetic model (Fig. 2).
Fig. 3. The failure mode of Model 3 showing the collapse of the medial plateau.
as the screws cut out through the bone. The failure load ranged from 1675 to 6096 N with a mean7SD of 376871482 N and a CV of 0.39. There was no significant difference (po0:05) in mean failure loads between the cadaver specimen and synthetic model 3.
3. Results In synthetic model 1 (solid-foam bone), failures were observed in one of two places (Fig. 2A). One was just distal to the end of the plates, and the other at the junction between the distal end of the tibial shaft and the mould. No failure occurred at the fracture site area covered by the fixation device. The mean failure load was 2200 (SD 78) N. In synthetic model 2 (composite bone), due to the strong mechanical properties of this cortical model, no collapse occurred in any specimen below 9000 N, which prevented us from detecting any failure of the fixation in near physiological forces that are clinically relevant. In synthetic model 3 (Reinforced solid-foam bone), only one failure mode was observed and it was a collapse of the medial plateau (Fig. 3). No failure occurred in the tibial shaft. The failure load ranged from 4150 to 4260 N with a mean (SD) of 4201 (44) N and a coefficient of variance (CV) of 0.01. All the cadaver specimens failed in the same mode as synthetic model 3, that is, the medial plateau collapsed
4. Discussion Loss of reduction is a common complication in the treatment of tibial plateau fractures, particularly in the elderly with osteoporosis. Biomechanical studies on cadaveric specimen have not identified which fixation method is the best for which type of the fracture, probably due to a wide variation in bone quality and test methods. We believe a standard synthetic bone model and test method can enhance the discriminate power of mechanical testing and facilitate meaningful comparison of different studies. In this study we aim to develop a bicondylar tibial plateau fracture model based on Sawbones and a testing method that reproduce the failure mode often seen in clinics. The use of a femoral component of a total knee replacement as a load applicator allowed us to load both condyles simultaneously, thus represents a more realistic loading condition compared with single condylar loading (Denny et al., 1984; Horwitz et al., 1999; Koval
ARTICLE IN PRESS A.M. Ali et al. / Journal of Biomechanics 39 (2006) 1355–1360
et al., 1996; Parker et al., 1999). Another concerns of single condylar loading is that, since many fixation methods involves both plateaus, destruction of one plateau would certainly affect the strength of the other. However, we only measured the total load applied, not individual load applied to the medial and lateral condyles. We, like other investigators (Denny et al., 1984; Horwitz et al., 1999; Koval et al., 1996; Parker et al., 1999), believe that the joint contact force is mainly responsible for the tibial plateau subsidence and therefore made no attempt to include the muscle forces, although we recognize that muscles impart forces to the bone and fixation devices and the forces change depending on the activities (Duda et al., 2001; Heller et al., 2001; Heller et al., 2003). Our plateau fracture model based on Sawbones’ 3rd generation composite tibia was almost un-destructible and withstood a compressive joint contact force over 12 times body weight of a typical 70 kg man. Although these Sawboness tibiae have been shown to have similar material (Currey, 1984; Sawbones Europe AB., 2002) and structural properties (Cristofolini and Viceconti, 2000; Sawbones Europe AB., 2002) to that of human bone, a ‘‘cortical’’ shell of about 3 mm covers all metaphyseal and articular surfaces and ‘‘cancellous bone’’ fills the whole proximal metaphysis, which may lead to unrealistically high local strength and explain our findings. Model 1 using Sawbones solid-form tibia failed at one of two sites some distance away from the fracture site (Fig. 2A). These failure modes reflected the material and structural weakness of the diaphysis. According to the Sawbones’ supplier the solid-foam used to construct Type 1116 tibia has a density of 0.32 g cm3, compressive Young’s modulus of 260 MPa and compressive strength of 8 MPa, which lies in the middle of the values for proximal tibial trabecular bone (Keyak et al., 1994). Comparing with human tibia, such construct may have similar mechanical properties in the metaphysis, but
1359
certainly is much weaker and flexible in the diaphysis (Table 1). The failure modes also reflected the testing conditions. It was observed that as loading of the specimens increased to around 2000 N, and in spite of the engagement of the sharp femoral spikes in the tibial surface, the tibia tended to slide anteriorly due to the nature of the posterior slope of the plateau, resulting in a bending moment in the sagittal plane. There were two stress risers: one proximally at the junction of the tibial shaft with the fixation implant, and the second is distally at the junction with the holding mould, causing the failure at those places. These failure modes may be avoided by fixing the specimen much more proximally, but since we planed to compare internal fixation method with external fixators, we had to leave enough space proximally for external fixators. The reinforced solid-foam bone model was the most reliable model for mechanical testing of bi-condylar tibial plateau fractures. Using this model, the only mode of failure was a collapse of the medial plateau. The mode of failure is not only identical to that of the cadaver model tested, but also similar to that reported clinically in many published studies (Benirschke et al., 1992; Bowes and Hohl, 1982; Delamarter et al., 1990; Marsh et al., 1995). Furthermore, the failure load of this model is within the range of the cadaver model. A simplified assessment of transverse-sectional bending stiffness of distal tibia diaphysis of different models (Table 1) showed that this model was an order of magnitude more rigid structurally than its un-reinforced counterpart Model 1, and as a result its diaphysis was able to withstand the applied load that caused the plateau subsidence and fixation failure, although, due to the absence of cortical bone, its bending stiffness was still much lower than the human tibia. It was noted that, contrast to the cadaver model, the failure load of Model 3 had a very small coefficient of variance. This will have a significant implication in the number of specimens required in order to obtain a
Table 1 Transverse-sectional area (A), 2nd moment of inertia (I) and bending stiffness (EI) of distal tibial diaphysis Types of bone
Geometrical Components
L (mm)
A (mm2)
I (mm4)
E (GPa)
EI (Nm2)
Human Tibiaa,b
a: Solid cross-section b: Medullary canal a–b: Cortical Shell
25 18–11
271 141–54 130–217
7047 1924–278 5123–6769
15–20 0 15–20
106.8–140.9
Model 1
a: Solid cross–section b: Medullary canal a–b: Model 1
25 6.35
271 127 144
7047 1276 5771
0.26 0 0.26
1.8 0 1.5
Model 3a
a: Model 1 b: Glass-fibre rod a+b: Model 3
144 127 271
5770 1276 7046
0.26 12.40
1.5 15.8 17.3
a
6.35
Note: a The tibial cross-section is assumed to be a 25 mm equal-sided triangle. b The cortical shell thickness of human tibia is assumed to be 3–6 mm and the Young’s modulus to be 15–20 GPa.
76.8–135.4
ARTICLE IN PRESS 1360
A.M. Ali et al. / Journal of Biomechanics 39 (2006) 1355–1360
statistically significant results. For example, with a standard deviation of 1482 N, at least 4000 specimens are required in order to detect a 75 N difference in the failure load with a 90% statistical power. In contrast to cadaver model, only 4 specimens of synthetic model 3 are required for the same 75 N differences in failure load to be detected with the same 90% statistical power. Failure of fixation of tibial plateau fractures is often found in the elderly with poor bone quality (Ali et al., 2002). We accept that the Sawbones model may not represent osteoporotic bone and the measured strength value may not be applicable to the elderly patients, but our model does allow a fair comparison of different fixation methods and can be adopted in synthetic tibia simulating osteoporosis once they become available. We have used this model to test 5 different methods of internal and external fixation successfully (Ali et al., 2003), and recommend it for future mechanical tests of the fixation of bi-condylar tibial plateau fractures.
Acknowledgements This work is supported by a grant from the Wishbone Trust. References Ali, A.M., El Shafie, M., Willett, K.M., 2002. Failure of fixation of tibial plateau fractures. Journal of Orthopaedic Trauma 16, 323–329. Ali, A.M., Saleh, M., Bolongaro, S., Yang, L., 2003. The strength of different fixation techniques for bicondylar tibial plateau fracturesa biomechanical study. Clinical Biomechanics 18, 864–870. Benirschke, S.K., Agnew, S.G., Mayo, K.A., Santoro, V.M., Henley, M.B., 1992. Immediate internal fixation of open, complex tibial plateau fractures: treatment by a standard protocol. Journal of Orthopaedic Trauma 6, 78–86. Blokker, C.P., Rorabeck, C.H., Bourne, R.B., 1984. Tibial plateau fractures. An analysis of the results of treatment in 60 patients. Clinical Orthopaedics and Related Research 182, 193–199. Bowes, D.N., Hohl, M., 1982. Tibial condylar fractures. Evaluation of treatment and outcome. Clinical Orthopaedics 171, 104–108. Burri, C., Bartzke, G., Coldewey, J., Muggler, E., 1979. Fractures of the tibial plateau. Clinical Orthopaedics 138, 84–93. Cristofolini, L., Viceconti, M., 2000. Mechanical validation of whole bone composite tibia models. Journal of Biomechanics 33, 279–288. Currey, J., 1984. The Mechanical Adaptations of Bones. Princeton University Press, Guildford. Delamarter, R.B., Hohl, M., Hopp Jr., E., 1990. Ligament injuries associated with tibial plateau fractures. Clinical Orthopaedics 250, 226–233. Denny, L., Keating, E., Engelhardt, J., Saha, S., 1984. A comparison of fixation techniques in tibial plateau fractures. 30th Annual ORS, Atlanta, Georgia, Febuary7–9, p 314. Duda, G.N., Mandruzzato, F., Heller, M., Goldhahn, J., Moser, R., Hehli, M., Claes, L., Haas, N.P., 2001. Mechanical boundary
conditions of fracture healing: borderline indications in the treatment of unreamed tibial nailing. Journal of Biomechanics 34, 639–650. Earnshaw, S.A., 2000. Biomechanical comparison of fixation of type-I fractures of the lateral tibial plateau. Journal of Bone and Joint Surgery 82B, 152–153. Einhorn, T.A., 1992. Bone strength: the bottom line. Calcified Tissue International 51, 333–339. Fernandez, D.L., 1988. Anterior approach to the knee with osteotomy of the tibial tubercle for bicondylar tibial fractures. Journal of Bone and Joint Surgery 70A, 208–219. Gerber, A., Ganz, R., 1998. Combined internal and external osteosynthesis a biological approach to the treatment of complex fractures of the proximal tibia. Injury 29 (Suppl 3), C22–C28. Goldstein, S.A., 1987. The mechanical properties of trabecular bone: dependence on anatomic location and function. Journal of Biomechanics 20, 1055–1061. Heller, M.O., Bergmann, G., Deuretzbacher, G., Durselen, L., Pohl, M., Claes, L., Haas, N.P., Duda, G.N., 2001. Musculo-skeletal loading conditions at the hip during walking and stair climbing. Journal of Biomechanics 34, 883–893. Heller, M.O., Taylor, W.R., Perka, C., Duda, G.N., 2003. The influence of alignment on the musculo-skeletal loading conditions at the knee. Langenbecks Archives of Surgery 388, 291–297. Hohl, M., 1967. Tibial condylar fractures. Journal of Bone and Joint Surgery 49A, 1455–1467. Horwitz, D.S., Bachus, K.N., Craig, M.A., Peters, C.L., 1999. A biomechanical analysis of internal fixation of complex tibial plateau fractures. Journal of Orthopaedic Trauma 13, 545–549. Keyak, J.H., Lee, I.Y., Skinner, H.B., 1994. Correlations between orthogonal mechanical properties and density of trabecular bone: use of different densitometric measures. Journal of Biomedical Materials. Research 28, 1329–1336. Koval, K.J., Polatsch, D., Kummer, F.J., Cheng, D., Zuckerman, J.D., 1996. Split fractures of the lateral tibial plateau: evaluation of three fixation methods. Journal of Orthopaedic Trauma 10, 304–308. Marsh, J.L., Smith, S.T., Do, T.T., 1995. External fixation and limited internal fixation for complex fractures of the tibial plateau. Journal of Bone and Joint Surgery 77A, 661–673. Parker, P.J., Tepper, K.B., Brumback, R.J., Novak, V.P., Belkoff, S.M., 1999. Biomechanical comparison of fixation of type-I fractures of the lateral tibial plateau. Is the antiglide screw effective? Journal of Bone and Joint Surgery 81B, 478–480. Prayson, M.J., Datta, D.K., Marshall, M.P., 2001. Mechanical comparison of endosteal substitution and lateral plate fixation in supracondylar fractures of the femur. Journal of Orthopaedic Trauma 15, 96–100. Roberts, J.M., 1968. Fractures of the condyles of the tibia. An anatomical and clinical end-result study of one hundred cases. Journal of Bone and Joint Surgery 50A, 1505–1521. Sawbones Europe AB., 2002. Product Catalog 2002. Schatzker, J., 1987. The rationale of operative fracture care. Springer, Berlin, pp. 279–295. Schatzker, J., McBroom, R., Bruce, D., 1979. The tibial plateau fracture. The Toronto experience 1968–1975. Clinical Orthopaedics 138, 94–104. Sievanen, H., Kannus, P., Oja, P., Vuori, I., 1994. Dual energy X-ray absorptiometry is also an accurate and precise method to measure the dimensions of human long bones. Calcified Tissue International 54, 101–105. Waddell, J.P., Johnston, D.W., Neidre, A., 1981. Fractures of the tibial plateau: a review of ninety-five patients and comparison of treatment methods. Journal of Trauma 21, 376–381.
Journal of Biomechanics 39 (2006) 1355–1360
Short communication
www.elsevier.com/locate/jbiomech www.JBiomech.com
Experimental model of tibial plateau fracture for biomechanical testing Ahmad M. Alia, Michael Salehb, Stefano Bolongarob, Lang Yangb, a
Department of Orthopedics and traumatology, Oldchurch Hospital, Waterloo Road, Romford, Essex RM7 0BE, UK Orthopaedic and Traumatic Surgery Research Group, Division of Clinical Sciences (North), University of Sheffield, Northern General Hospital, Sheffield S5 7AU, UK
b
Accepted 11 March 2005
Abstract Although adequate reduction and stable fixation have been recognized to be the prime goals in the treatment of displaced tibial plateau fractures, the optimal fixation technique remains controversial. The lack of a reliable model and a standard methodology contribute to this situation. The purpose of this study is to develop an experimental model of a tibial plateau fracture and a testing methodology that reproduces the failure mode commonly seen in the clinical setting. Using solid-foam and composite Sawbones tibiae, three different models of bi-condylar tibial plateau fracture (solid-foam, reinforced solid-foam and composite), six specimens for each model, were created and stabilized with double plating. The specimens were subjected to cyclic axial compression with increasing maximum load until failure. A femoral component of a total knee replacement of similar size and shape to the synthetic tibial surface was used as a load applicator. The experiment was repeated on six specimens of human cadaver tibiae. Among the Sawbones specimens, only the reinforced solid-foam model was found to produce a consistent failure mode (collapse in the medial plateau) comparable to that reported clinically in the literature. This mode of failure was also confirmed by the cadaver experiments. The failure load of the reinforced solid-foam model ranged from 4150 to 4260 N with a mean7SD of 4201744 N and a coefficient of variance of 0.01, whereas for the cadaver model the failure load ranged from 1675 to 6096 N with a mean7SD of 376871482 N and a coefficient of variance of 0.39. We recommend the reinforced-foam model for future mechanical tests to compare different fixation methods for tibial plateau fractures. r 2005 Elsevier Ltd. All rights reserved. Keywords: Biomechanical testing; Experimental model; Tibial plateau fracture; Sawbones
1. Introduction Tibial plateau fracture is a common injury, with an incidence of 9.2% of all tibial fractures (Burri et al., 1979; Fernandez, 1988; Hohl, 1967; Roberts, 1968). Although adequate reduction and stable fixation have been recognized to be the primary goals in the treatment of displaced tibial plateau fractures, the optimal fixation technique remains controversial (Blokker et al., 1984; Hohl, 1967; Schatzker et al., 1979; Waddell et al., 1981). Corresponding author. Tel: +44 114 2714900; fax: +44 114 2619246. E-mail address: l.yang@sheffield.ac.uk (L. Yang).
0021-9290/$ - see front matter r 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbiomech.2005.03.022
Loss of reduction is frequent and may depend on the type of fixation (Benirschke et al., 1992; Bowes and Hohl, 1982; Delamarter et al., 1990; Hohl, 1967; Marsh et al., 1995). The biomechanical data based on cadaveric bone have failed to show conclusively which fixation method is best for each type of tibial plateau fracture (Denny et al., 1984; Horwitz et al., 1999; Koval et al., 1996; Parker et al., 1999). Furthermore, conflicting conclusions were reached in two different studies which tested the same type of fracture using the same fixation technique (Denny et al., 1984; Koval et al., 1996). The use of cadavers for mechanical testing has merits and drawbacks. The obvious advantages are the anatomical and mechanical reality. The drawbacks are
ARTICLE IN PRESS 1356
A.M. Ali et al. / Journal of Biomechanics 39 (2006) 1355–1360
mainly the difficulties in obtaining a sufficient number of specimens (Earnshaw, 2000) (Einhorn, 1992; Goldstein, 1987) and the wide variability of bone quality (mechanical property) (Einhorn, 1992; Goldstein, 1987). The wide variability of bone quality leads to a wide variability in mechanical quantity testing, which in turn demands a greater number of specimen to ascertain a satisfactory statistical power. Without quantifying bone quality and random assignment, it is possible that one fixation method is tested mainly on specimens of poor bone quality and another on good quality bones. This may explain the differences found in the studies by Denny and Koval (Denny et al., 1984; Koval et al., 1996). Bone mineral density can be measured readily by dual energy x-ray absorptiometry (DXA), but the density on its own is not totally representative of bone quality (Sievanen et al., 1994). Other variables, like orientation, thickness and separation of the trabeculae are also important factors that determine the mechanical properties of cancellous bone (Einhorn, 1992; Goldstein, 1987). Unfortunately, these parameters are much more difficult to quantify. With a limited number of specimens and without proper control of bone quality between the comparative groups, the validity of the conclusions from previous studies is questionable (Earnshaw, 2000). The method used in biomechanical testing is another important variable that makes it difficult to compare the results of different studies. Parker et al. (1999) used a polyurethane load applicator to ramp compress the lateral plateau, and the yield force was defined as the force at 2 mm plateau subsidence. Horwitz et al. (1999) and Koval et al. (1996) used a steel ball as load applicator, ramp and cyclically loaded the specimen, and the total subsidence between fixation methods were compared. There is a need for a reliable model of tibial plateau fracture for mechanical testing, as well as a standard methodology permitting comparison to be made between different studies and of different fixation techniques. Synthetic bone material has been used previously in the mechanical testing of fixation devices (Gerber and Ganz, 1998; Prayson et al., 2001). Sawboness limb models (Sawbones Europe AB, Sweden) have been shown to have similar material and structural properties to that of human bone (Cristofolini and Viceconti, 2000; Sawbones Europe AB., 2002). The obvious advantages of using Sawboness are the assurance of uniformity, consistency in its material and geometry, and therefore its mechanical properties. This synthetic bone model has to be compared to a cadaveric model to validate the mode of failure. With such a model, the performance of fixation devices can be evaluated more reliably since the variability within specimens is properly controlled. As the first phase of an investigation into the effects of bone quality on the strength of different fixation
methods for tibial plateau fracture, the aim of this study is to develop a model of tibial plateau fracture and a testing methodology that consistently produces the failure mode commonly seen in clinical settings.
2. Materials and methods Using Sawboness tibia three different models were employed in the first part of this study (Fig. 1): Model 1—solid-foam bone (Sawboness Product no. 1116): a rigid polyurethane foam material throughout that simulates cancellous bone. It has a circular ‘‘medullary canal’’ of 12 mm in diameter. Model 2—composite bone (Sawboness Product no. 3301): the core of the bone is made of rigid polyurethane foam and the cortex layer is formed by pressure injecting a mixture of short e-glass fibres and epoxy resin around the form core. The structural properties of the whole tibial model have been validated (Cristofolini and Viceconti, 2000). Model 3—reinforced solid-foam bone: the same as the first model but reinforced with an ‘‘intramedullary’’ carbon-fibre rod (Tufnols 10G/40, Tuflon Ltd, Birmingham, UK) to support the diaphysis. The rod was 12.7 mm (0.5 in) in diameter and inserted into the tibial canal about 85 mm distal to the plateau and just distal to the most distal screws of the fixation plate. This model was created specifically for this study. The mechanical properties of the Sawbones materials were found to be in the range of natural bones (Currey, 1984; Sawbones Europe AB., 2002). The Sawbones were obtained from one manufacturing batch ensuring the same material properties and geometry. Six specimens of each model were used. A bi-condylar tibial plateau fracture (AO/OTA type 41-C1, Schatzker type-V (Schatzker et al., 1979)) was created using an automated thin blade saw (Fig. 1A). A template was used to create a reproducible cut with the same inclination and distances from the tibial spine to the lateral and medial edges of the tibial plateau. The fractures were reduced and fixed using the AO dual plating technique, this was performed by a single orthopaedic surgeon (AMA) who was fully trained in the AO principles of internal fixation (Schatzker, 1987) (Fig. 2). This dual plating technique involved two L shaped buttress plates (six hole), one for each condylar fracture, with three partially threaded 6.5 mm cancellous screws proximally and three 4.5 mm cortical screws distally in each plate. The model was fixed in an engineering angle vice, via a split resin mould that held the distal tibia rigidly, with the long axis of the tibial shaft at 31 varus and a 101 posterior slope to simulate the normal alignment in midstance full weight bearing (Fig. 2). The specimen was then mounted and fixed onto a material-testing machine (Autograph ASG10kN, Shimadzu Corporation, Kyoto,
ARTICLE IN PRESS A.M. Ali et al. / Journal of Biomechanics 39 (2006) 1355–1360
1357
Fig. 1. (A) Bi-condylar tibial plateau fracture (Schatzker Type-V). (B) Model 1: solid-foam bone. (C) Model 2: composite bone. (D) Model 3: reinforced solid-foam bone with an intramedullary carbon-fibre rod (pointed by an arrow) to support the diaphysis.
Fig. 2. (A) Synthetic and (B) cadaver bi-condylar tibial plateau fracture models instrumented and ready for mechanical testing. The arrows point to the sites of the stress riser where the failure occurred in Model 1.
Japan). In order to measure the collapse of the two plateaus under loading, four extensometers, two for each plateau, were mounted on the model anteromedially, postero-medially, antero-laterally and pos-
tero-laterally (Fig. 2). These extensometers were calibrated against the crosshead movement (0.001 mm resolution) and found to be accurate within 5%. In the process of developing a suitable loading applicator, different applicators described in other studies were tried (Horwitz et al., 1999; Koval et al., 1996; Parker et al., 1999). However, these applicators were discarded as they were found to cause a focal crushing effect localized to one area, instead of collapse in the plateau fixation. Furthermore, some of these applicators (spherical ball) are designed to test only one plateau. Therefore, a femoral component of a total knee replacement of a similar size and shape to the tibial surface was finally selected and used (Size 4 PFCs SigmaTM Keen System, DePuy Orthopaedics Inc., Warsaw, USA) as a loading applicator in order to deliver forces on both tibial plateau surfaces simultaneously (Fig. 2). Initial testing revealed that as loading increased to 2000 N, a consistent posterior slide or translation of the femur on the tibial articular surface occurred. This was due to the posterior tibial slope and the low friction between the two surfaces. To avoid this sliding, we inserted two small sharp spikes (2 mm) on the articular surface of the femoral prosthesis, one in each condyle to engage the tibial surface during testing (Fig. 2). This proved to be effective in preventing any sliding or subluxation during axial loading. Cyclical compression was applied to the specimen at a rate of 20 mm/min. For conditioning the system four
ARTICLE IN PRESS 1358
A.M. Ali et al. / Journal of Biomechanics 39 (2006) 1355–1360
loading cycles at 500 N were first performed. The actual tests started with four loading cycles at 1000 N, followed by four loading cycles at 2000 N, four cycles at 4000 N and four cycles at 9000 N, if there was no obvious failure of the construct during the previous loading level, otherwise the experiment stopped at the load level where an obvious failure of the construct occurred. The output from the material testing machine (load and crosshead displacement) and extensometers were recorded for later analysis at 10 Hz using a computerized data acquisition system. Failure was defined as a plastic displacement of 3 mm or more in the articular surface in either condyle. A previous study has shown that 3 mm represented a significant step-off displacement, which is unacceptable clinically (Ali et al., 2002). In the second part of the study, six human cadaveric tibiae of similar size and bone density were used. The tibiae were collected from subjects without a medical history of skeletal pathology (mean7SD age ¼ 66.57 10 years, range 53–78 years). Bi-planar radiographs were used to exclude those specimens with other pathological lesions. Following soft tissue removal, they were wrapped in saline soaked gauze sponges and stored in plastic bags at 70 1C. They were thawed in room temperature before experiment. DXA scanning using Hologic QDR 4500 was performed on all specimens, and it confirmed that these tibiae are a homogenous group with an average BMD of 0.65 g/cm2 (range 0.55–0.75 g/cm2; SD 0.08 g/cm2). The method of creating and stabilizing the fracture and the protocol of mechanical testing was identical to that used for the synthetic model (Fig. 2).
Fig. 3. The failure mode of Model 3 showing the collapse of the medial plateau.
as the screws cut out through the bone. The failure load ranged from 1675 to 6096 N with a mean7SD of 376871482 N and a CV of 0.39. There was no significant difference (po0:05) in mean failure loads between the cadaver specimen and synthetic model 3.
3. Results In synthetic model 1 (solid-foam bone), failures were observed in one of two places (Fig. 2A). One was just distal to the end of the plates, and the other at the junction between the distal end of the tibial shaft and the mould. No failure occurred at the fracture site area covered by the fixation device. The mean failure load was 2200 (SD 78) N. In synthetic model 2 (composite bone), due to the strong mechanical properties of this cortical model, no collapse occurred in any specimen below 9000 N, which prevented us from detecting any failure of the fixation in near physiological forces that are clinically relevant. In synthetic model 3 (Reinforced solid-foam bone), only one failure mode was observed and it was a collapse of the medial plateau (Fig. 3). No failure occurred in the tibial shaft. The failure load ranged from 4150 to 4260 N with a mean (SD) of 4201 (44) N and a coefficient of variance (CV) of 0.01. All the cadaver specimens failed in the same mode as synthetic model 3, that is, the medial plateau collapsed
4. Discussion Loss of reduction is a common complication in the treatment of tibial plateau fractures, particularly in the elderly with osteoporosis. Biomechanical studies on cadaveric specimen have not identified which fixation method is the best for which type of the fracture, probably due to a wide variation in bone quality and test methods. We believe a standard synthetic bone model and test method can enhance the discriminate power of mechanical testing and facilitate meaningful comparison of different studies. In this study we aim to develop a bicondylar tibial plateau fracture model based on Sawbones and a testing method that reproduce the failure mode often seen in clinics. The use of a femoral component of a total knee replacement as a load applicator allowed us to load both condyles simultaneously, thus represents a more realistic loading condition compared with single condylar loading (Denny et al., 1984; Horwitz et al., 1999; Koval
ARTICLE IN PRESS A.M. Ali et al. / Journal of Biomechanics 39 (2006) 1355–1360
et al., 1996; Parker et al., 1999). Another concerns of single condylar loading is that, since many fixation methods involves both plateaus, destruction of one plateau would certainly affect the strength of the other. However, we only measured the total load applied, not individual load applied to the medial and lateral condyles. We, like other investigators (Denny et al., 1984; Horwitz et al., 1999; Koval et al., 1996; Parker et al., 1999), believe that the joint contact force is mainly responsible for the tibial plateau subsidence and therefore made no attempt to include the muscle forces, although we recognize that muscles impart forces to the bone and fixation devices and the forces change depending on the activities (Duda et al., 2001; Heller et al., 2001; Heller et al., 2003). Our plateau fracture model based on Sawbones’ 3rd generation composite tibia was almost un-destructible and withstood a compressive joint contact force over 12 times body weight of a typical 70 kg man. Although these Sawboness tibiae have been shown to have similar material (Currey, 1984; Sawbones Europe AB., 2002) and structural properties (Cristofolini and Viceconti, 2000; Sawbones Europe AB., 2002) to that of human bone, a ‘‘cortical’’ shell of about 3 mm covers all metaphyseal and articular surfaces and ‘‘cancellous bone’’ fills the whole proximal metaphysis, which may lead to unrealistically high local strength and explain our findings. Model 1 using Sawbones solid-form tibia failed at one of two sites some distance away from the fracture site (Fig. 2A). These failure modes reflected the material and structural weakness of the diaphysis. According to the Sawbones’ supplier the solid-foam used to construct Type 1116 tibia has a density of 0.32 g cm3, compressive Young’s modulus of 260 MPa and compressive strength of 8 MPa, which lies in the middle of the values for proximal tibial trabecular bone (Keyak et al., 1994). Comparing with human tibia, such construct may have similar mechanical properties in the metaphysis, but
1359
certainly is much weaker and flexible in the diaphysis (Table 1). The failure modes also reflected the testing conditions. It was observed that as loading of the specimens increased to around 2000 N, and in spite of the engagement of the sharp femoral spikes in the tibial surface, the tibia tended to slide anteriorly due to the nature of the posterior slope of the plateau, resulting in a bending moment in the sagittal plane. There were two stress risers: one proximally at the junction of the tibial shaft with the fixation implant, and the second is distally at the junction with the holding mould, causing the failure at those places. These failure modes may be avoided by fixing the specimen much more proximally, but since we planed to compare internal fixation method with external fixators, we had to leave enough space proximally for external fixators. The reinforced solid-foam bone model was the most reliable model for mechanical testing of bi-condylar tibial plateau fractures. Using this model, the only mode of failure was a collapse of the medial plateau. The mode of failure is not only identical to that of the cadaver model tested, but also similar to that reported clinically in many published studies (Benirschke et al., 1992; Bowes and Hohl, 1982; Delamarter et al., 1990; Marsh et al., 1995). Furthermore, the failure load of this model is within the range of the cadaver model. A simplified assessment of transverse-sectional bending stiffness of distal tibia diaphysis of different models (Table 1) showed that this model was an order of magnitude more rigid structurally than its un-reinforced counterpart Model 1, and as a result its diaphysis was able to withstand the applied load that caused the plateau subsidence and fixation failure, although, due to the absence of cortical bone, its bending stiffness was still much lower than the human tibia. It was noted that, contrast to the cadaver model, the failure load of Model 3 had a very small coefficient of variance. This will have a significant implication in the number of specimens required in order to obtain a
Table 1 Transverse-sectional area (A), 2nd moment of inertia (I) and bending stiffness (EI) of distal tibial diaphysis Types of bone
Geometrical Components
L (mm)
A (mm2)
I (mm4)
E (GPa)
EI (Nm2)
Human Tibiaa,b
a: Solid cross-section b: Medullary canal a–b: Cortical Shell
25 18–11
271 141–54 130–217
7047 1924–278 5123–6769
15–20 0 15–20
106.8–140.9
Model 1
a: Solid cross–section b: Medullary canal a–b: Model 1
25 6.35
271 127 144
7047 1276 5771
0.26 0 0.26
1.8 0 1.5
Model 3a
a: Model 1 b: Glass-fibre rod a+b: Model 3
144 127 271
5770 1276 7046
0.26 12.40
1.5 15.8 17.3
a
6.35
Note: a The tibial cross-section is assumed to be a 25 mm equal-sided triangle. b The cortical shell thickness of human tibia is assumed to be 3–6 mm and the Young’s modulus to be 15–20 GPa.
76.8–135.4
ARTICLE IN PRESS 1360
A.M. Ali et al. / Journal of Biomechanics 39 (2006) 1355–1360
statistically significant results. For example, with a standard deviation of 1482 N, at least 4000 specimens are required in order to detect a 75 N difference in the failure load with a 90% statistical power. In contrast to cadaver model, only 4 specimens of synthetic model 3 are required for the same 75 N differences in failure load to be detected with the same 90% statistical power. Failure of fixation of tibial plateau fractures is often found in the elderly with poor bone quality (Ali et al., 2002). We accept that the Sawbones model may not represent osteoporotic bone and the measured strength value may not be applicable to the elderly patients, but our model does allow a fair comparison of different fixation methods and can be adopted in synthetic tibia simulating osteoporosis once they become available. We have used this model to test 5 different methods of internal and external fixation successfully (Ali et al., 2003), and recommend it for future mechanical tests of the fixation of bi-condylar tibial plateau fractures.
Acknowledgements This work is supported by a grant from the Wishbone Trust. References Ali, A.M., El Shafie, M., Willett, K.M., 2002. Failure of fixation of tibial plateau fractures. Journal of Orthopaedic Trauma 16, 323–329. Ali, A.M., Saleh, M., Bolongaro, S., Yang, L., 2003. The strength of different fixation techniques for bicondylar tibial plateau fracturesa biomechanical study. Clinical Biomechanics 18, 864–870. Benirschke, S.K., Agnew, S.G., Mayo, K.A., Santoro, V.M., Henley, M.B., 1992. Immediate internal fixation of open, complex tibial plateau fractures: treatment by a standard protocol. Journal of Orthopaedic Trauma 6, 78–86. Blokker, C.P., Rorabeck, C.H., Bourne, R.B., 1984. Tibial plateau fractures. An analysis of the results of treatment in 60 patients. Clinical Orthopaedics and Related Research 182, 193–199. Bowes, D.N., Hohl, M., 1982. Tibial condylar fractures. Evaluation of treatment and outcome. Clinical Orthopaedics 171, 104–108. Burri, C., Bartzke, G., Coldewey, J., Muggler, E., 1979. Fractures of the tibial plateau. Clinical Orthopaedics 138, 84–93. Cristofolini, L., Viceconti, M., 2000. Mechanical validation of whole bone composite tibia models. Journal of Biomechanics 33, 279–288. Currey, J., 1984. The Mechanical Adaptations of Bones. Princeton University Press, Guildford. Delamarter, R.B., Hohl, M., Hopp Jr., E., 1990. Ligament injuries associated with tibial plateau fractures. Clinical Orthopaedics 250, 226–233. Denny, L., Keating, E., Engelhardt, J., Saha, S., 1984. A comparison of fixation techniques in tibial plateau fractures. 30th Annual ORS, Atlanta, Georgia, Febuary7–9, p 314. Duda, G.N., Mandruzzato, F., Heller, M., Goldhahn, J., Moser, R., Hehli, M., Claes, L., Haas, N.P., 2001. Mechanical boundary
conditions of fracture healing: borderline indications in the treatment of unreamed tibial nailing. Journal of Biomechanics 34, 639–650. Earnshaw, S.A., 2000. Biomechanical comparison of fixation of type-I fractures of the lateral tibial plateau. Journal of Bone and Joint Surgery 82B, 152–153. Einhorn, T.A., 1992. Bone strength: the bottom line. Calcified Tissue International 51, 333–339. Fernandez, D.L., 1988. Anterior approach to the knee with osteotomy of the tibial tubercle for bicondylar tibial fractures. Journal of Bone and Joint Surgery 70A, 208–219. Gerber, A., Ganz, R., 1998. Combined internal and external osteosynthesis a biological approach to the treatment of complex fractures of the proximal tibia. Injury 29 (Suppl 3), C22–C28. Goldstein, S.A., 1987. The mechanical properties of trabecular bone: dependence on anatomic location and function. Journal of Biomechanics 20, 1055–1061. Heller, M.O., Bergmann, G., Deuretzbacher, G., Durselen, L., Pohl, M., Claes, L., Haas, N.P., Duda, G.N., 2001. Musculo-skeletal loading conditions at the hip during walking and stair climbing. Journal of Biomechanics 34, 883–893. Heller, M.O., Taylor, W.R., Perka, C., Duda, G.N., 2003. The influence of alignment on the musculo-skeletal loading conditions at the knee. Langenbecks Archives of Surgery 388, 291–297. Hohl, M., 1967. Tibial condylar fractures. Journal of Bone and Joint Surgery 49A, 1455–1467. Horwitz, D.S., Bachus, K.N., Craig, M.A., Peters, C.L., 1999. A biomechanical analysis of internal fixation of complex tibial plateau fractures. Journal of Orthopaedic Trauma 13, 545–549. Keyak, J.H., Lee, I.Y., Skinner, H.B., 1994. Correlations between orthogonal mechanical properties and density of trabecular bone: use of different densitometric measures. Journal of Biomedical Materials. Research 28, 1329–1336. Koval, K.J., Polatsch, D., Kummer, F.J., Cheng, D., Zuckerman, J.D., 1996. Split fractures of the lateral tibial plateau: evaluation of three fixation methods. Journal of Orthopaedic Trauma 10, 304–308. Marsh, J.L., Smith, S.T., Do, T.T., 1995. External fixation and limited internal fixation for complex fractures of the tibial plateau. Journal of Bone and Joint Surgery 77A, 661–673. Parker, P.J., Tepper, K.B., Brumback, R.J., Novak, V.P., Belkoff, S.M., 1999. Biomechanical comparison of fixation of type-I fractures of the lateral tibial plateau. Is the antiglide screw effective? Journal of Bone and Joint Surgery 81B, 478–480. Prayson, M.J., Datta, D.K., Marshall, M.P., 2001. Mechanical comparison of endosteal substitution and lateral plate fixation in supracondylar fractures of the femur. Journal of Orthopaedic Trauma 15, 96–100. Roberts, J.M., 1968. Fractures of the condyles of the tibia. An anatomical and clinical end-result study of one hundred cases. Journal of Bone and Joint Surgery 50A, 1505–1521. Sawbones Europe AB., 2002. Product Catalog 2002. Schatzker, J., 1987. The rationale of operative fracture care. Springer, Berlin, pp. 279–295. Schatzker, J., McBroom, R., Bruce, D., 1979. The tibial plateau fracture. The Toronto experience 1968–1975. Clinical Orthopaedics 138, 94–104. Sievanen, H., Kannus, P., Oja, P., Vuori, I., 1994. Dual energy X-ray absorptiometry is also an accurate and precise method to measure the dimensions of human long bones. Calcified Tissue International 54, 101–105. Waddell, J.P., Johnston, D.W., Neidre, A., 1981. Fractures of the tibial plateau: a review of ninety-five patients and comparison of treatment methods. Journal of Trauma 21, 376–381.