Research Article

Hexapod Fixator Mechanics - The Final Link

Warren Macdonald¹, Ling Feng Zhang¹, Man To Yeung¹, Jedidiah Farquharson¹ and Satyajit Naique²

¹Department of Bioengineering, Imperial College London, Prince Consort Road, South Kensington, London, UK

²Trauma and Orthopaedic Surgery, St Mary’s Hospital, Paddington, Imperial College NHS Healthcare Trust, London, UK

Submission:September 09, 2024; Published:October 17, 2024

*Corresponding author: Dr Warren Macdonald, Department of Bioengineering, Imperial College London, Prince Consort Road, South Kensington, London, SW7 2BP, UK

How to cite this article:Warren M, Ling Feng Z, Man To Y, Jedidiah F, Satyajit N. Hexapod Fixator Mechanics - The Final Link. Ortho & Rheum Open Access J. 2024; 23(5): 556125.DOI: 10.19080/OROAJ.2024.23.556125

Abstract

Background: Hexapod External fixators are used for complex limb reconstruction, limb lengthening and deformity correction: recently they have also provided definitive fixation for complex fractures. An accurate knowledge of hexapod biomechanics will enable the clinician to unleash full capabilities of the device, especially its control of the biomechanical environment at the osteotomy /fracture site, creating an optimal mechanical strain with minimal biological insult. Recent studies have extensively characterized the Taylor Spatial Frame (TSF, Richards, USA) system but only some aspects of the newer Trulok Hex System (TL-HEX; Orthofix SRL Italy).
Methods: Standard frame constructs of TSF and TL-HEX were mechanically tested in axial tension/compression, transverse bending and axial torsion. Frames were mounted on bone analogues (fiber glass tubes) supported by paired crossed wires and the `Fixation’ constructs were tested mechanically.
Findings: The maximum axial stiffness of both unmounted systems was similar: 1552 N.mm^-1 and 1699 N.mm^-1. The overall mean stiffnesses were also similar: 302 vs 452 N.mm^-1. With crossed wires and bone analogues, the stiffness of fixation constructs dropped about 85% to around 26 N.mm^-1 for both: the stiffness of the hexapod was overshadowed by the low stiffness of the Fixation Elements and the deformation was non-linear.
Interpretation: Hexapod fixators are commonly used in clinical practice, but their complex biomechanics are not well understood. By clarifying this to clinicians, more informed choices can be made about the fixation mechanics and hence the biomechanics of the fracture can be optimized to enhance fracture healing.

Keywords:Fracture Fixation; External Fixation; Hexapod fixators; Mechanical Testing; Stiffness

Introduction

Fracture healing in long bones depends on stable fixation of the bone fragments, in a controlled biological environment. For most fractures, this can be reliably achieved using internal fixation devices. However, for more complex Open fractures characterized by significant soft tissue injury, bone loss, wound contamination/infection and vascular compromise, external fixation is frequently used. This involves the use of percutaneous transfixion tensioned wires or half-pins fixed to the bone fragments and connected to the mechanical bracing structure, which can be planar (unilateral or bilateral [1]) or circular (Ilizarov type or hexapod) [2-4].

The mechanobiology of fracture healing is now well understood, and even in the most complex of fracture geometries and conformations it should be possible to define the ideal fixation stiffness and provide optimal bone-healing mechanics in the fracture zone [5-7]. A Circular External fixator, by virtue of its design and versatility, allows the surgeon limitless options for bone fixation and hence an ability to change the mechanical environment during bone healing (fracture/osteotomy). Moreover, it enables surgeons to make fine adjustments to the bone position and change the strain across the bone gap as required. This can promote bone healing using controlled micromotion at the fracture gap [6,8-10] in a clinical setting, without resorting to additional surgery.

The fine wire Circular Fixator as exemplified by the Ilizarov System ushered in a new era in reconstructive and trauma surgery, giving the surgeon flexibility in managing complex fractures, non-unions and bone deformities not available before [3,11]. The newer Hexapod devices are a modification of the traditional circular frames, using 6 telescoping rods connecting two reinforced rings (Figures 3 & 5); in a geometrical arrangement known in Mechanical Engineering as a Stewart-Gough platform [12,13]. They are technically easier to use in surgery and allow the use of a simplified web-based software program, which can be used in a clinic setting to manipulate the bone fragments (fracture or osteotomy), to achieve accurate corrections and realignment [4,14-17].

The biomechanics of the Taylor Spatial Frame system (TSF, Smith & Nephew Richards, Memphis, USA) hexapod frame and its fixation elements have been well documented [8,18-20]. For this hexapod, particular concern has been expressed about the “backlash” (Figure 4) which allows approximately 1 mm of motion at small loads (also known as “the rattle” [21]). This has been attributed to the use of Cardan joints in the strut connections [22]. By contrast, the TLHEX struts assemble into the rings through integral pivot pins inserted radially into matching sockets in the ring (Figure 3), which are then locked in place by a grub screw. This obviates “the rattle”, but the pins are cantilevered out from the rings and thus deflect on loading.

It is also noted that the fixation elements – Wires and half pins, demonstrate low stiffness (20% of the hexapod stiffness) initially but greater stiffness as the loads increase [23]. A functional external fixator is thus composed of several elements with different stiffnesses at the range of loads probably experienced. The biomechanics of the newer Trulok Hexapod System (TLHEX; Orthofix S.R.L., Bussolengo, Verona, Italy) have been reported [24], but this system has not been evaluated in clinically relevant constructs.

We have characterized the biomechanics of this system and report the meaning of these results in the context of fracture zone mechanics. Summing the importance of these various stiffnesses is complex and unnecessary – especially since the hexapod stiffness is non-linear and cannot be summarized by one single parameter. The important clinical parameter is the magnitude of strain experienced by the tissues in the fracture zone; which dictates the rate or extent of the fracture healing response. Indeed, different studies report stiffnesses in different ways and with different units [2,18,25]. A harmonizing approach would be extremely useful. Nevertheless, we have characterized the mechanics of the TSF and TLHEX frames and components, described by three parameters, and using Analysis of Variance have determined that the two designs are different and display different mechanics.

Methods

Components and assemblies of the Trulok Hexapod System (TLHEX; Orthofix S.R.L., Bussolengo, Verona, Italy) and Taylor Spatial Frame (TSF; Smith & Nephew Richards, Memphis, USA) fixator systems have been mechanically tested using a Materials Testing Machine (Instron 6686, Instron Ltd, High Wycombe, UK) and custom jigs. Individual struts (TL-Hex Medium) were connected into rings at either end, which were then clamped in the test machine grips for loading, thereby replicating the mechanical contribution of the individual strut to the overall construct mechanics (Figure 1). Struts were tested fully extended, half extended and fully retracted, under loads between +/- 600N; which is equivalent to Full Body weight for a 60 kg adult, a representative physiological load [19].

Loading was applied at a crosshead speed of 60 mm/min., automatically reversing when the extreme load had been reached; ten load cycles were applied in each test. Load and extension were sampled at 250 Hz and recorded. After each test, struts were repositioned in the grips and the test was repeated; five such tests were applied to each strut (Figure 2). A video camera was used to record video of the pin behavior under testing. Data from the tests was then analyzed to find a polynomial of best fit to all the cycles applied to each test piece. The mechanics were described by three parameters: an overall generalized stiffness, the maximum stiffness in compression and the maximum stiffness in tension. These three parameters, for both systems, were then subjected to an Analysis of Variance using MS Excel.

Frame constructs for both hexapod systems were assembled using 180 mm rings and six Medium struts, and then tested at three ring-to-ring separations (153, 159 and 161 mm), using similar strut extensions to those used in the individual strut tests. Due to slight differences in dimensions between the TL-HEX struts and those of the TSF system, similar ring-to ring distances required different strut extensions; matched strut extensions were then tested at 145, 162 and 180 mm strut length. Polyacetal discs of 20 mm thickness and 180 mm diameter were bolted behind the rings of the hexapod to connect to the Materials Testing Machine crosshead (Figure 3). The constructs were similarly loaded between +/- 600 N for 6 cycles (Figure 4).

Bone-fixator constructs were also assembled using bone analogues of glass-fiber reinforced polyester (GFRP) hollow tubes of 25 mm diameter and 3 mm wall thickness (grade RLG1, E= 7.1 GPa; Tufnol Ltd, Birmingham, UK), connected by paired crossed wires of 1.8 mm diameter, tensioned to 1100 N and locked to single rings of the hexapods (Figure 5). Since the Elastic Modulus of cortical bone varies between 21 GPa and 7 GPa [26], this material is a good analogue for cortical bone and long bones. Loading was applied to the ends of the bone analogue tubes, increasing from 10 to 200 N and then cycling to -200 N and back to 200N for 5 cycles (Figure 7), at a rate of 60 mm/min. This rate is “quasi-static” , so the viscoelasticity of bone is not required to be modelled. Constructs were also assembled with successive additional rings or half-rings and paired crossed wires or halfpins, and tested under the same conditions.

The bone-fixator constructs were also loaded in torsion and bending using specialized jigs. Torques of +/- 20 Nm were applied to the top end of the bone analogue of the construct whilst the distal “bone end” was clamped securely. Bending was applied using a four-point bending jig up to a moment of 58 Nm in one direction only. Deflections were measured in angular degrees in both torsion and bending (Figure 8). These loadings also cover the range of physiological loadings [19].

Results

Load deformation curves for the individual struts (Figure 2) demonstrated similar non-linear behavior for the TSF and TL-Hex systems. The TL-HEX struts demonstrated a low-stiffness region at initial loading, or when crossing from tension to compression or reverse; this was similar to that displayed by the TSF system. The struts’ behavior was also dependent on their conformation; whether extended or retracted.

When combined into Hexapod assemblies the systems demonstrated non-linear behavior, characterized by a lowstiffness region at initial loading (215 to 300 N.mm^-1), or when crossing from tension to compression or reverse (Figure 4). At higher loads the hexapod stiffnesses increased. The TSF stiffness ranged from 200 Nmm^-1 initially to 1699 Nmm^-1 at 1 kN load, whilst the TLHEX ranged from 300 Nmm^-1 initially to 1552 Nmm^{- 1} at 1 kN (Table 1). Overall (mean) stiffness for the two systems was also similar: 452 Nmm^-1 (TSF)302 N mm^-1 (TLHEX) The TSF demonstrated 0.8 mm slack or backlash between the linear stiffness regions, which ranged from 300 Nmm^-1 initially to 923 Nmm^-1 at maximum. The TL-Hex showed 0.6 mm motion between stiffnesses of 215 Nmm^-1 initially and 1200 Nmm^-1 max.

Analysis of Variance was significant for the differences between the two systems on each of the parameters (at p <0.01): the two systems behave differently at mean and peak stiffnesses (Table 4: Supplementary Material). Post-hoc pairwise comparisons of the two systems were then undertaken, showing statistically significant differences for each parameter (p < 0.025 in all cases) using the Bonferroni correction. By contrast, the overall stiffnesses found with bone/fixator constructs were about 85% lower (Figure 7), and the non-linear behavior was much less pronounced than with struts and hexapods alone. Indeed, the nonlinear behavior of the hexapods was masked by the less stiff linear behavior of the wires. With bone analogues, the TLHEX construct (with one ring and one pair of crossed wires) displayed a stiffness of 25-26 NNmm^-1 and backlash of 0.65 mm, whilst the TSF construct demonstrates stiffness of 27-29 Nmm^-1 and backlash of 2.1 mm (Table 2). Adding additional Bone Fixation Elements increased the construct stiffness linearly (Tables 5 & 6).

In torsion, our testing of the TLHEX/bone analogue also showed mildly non-linear behavior, with stiffness ranging from 0.5 Nm/° initially to around 2.5 Nm/° (Figure 8), which was consistent with the stiffnesses reported by Nikonovas [4].

Note 1: Torsional stiffness data for TSF from Nikonovas [23].

Discussion

To achieve stable fixation and enable optimal fracture healing, it is known that the tissue strain at the fracture gap is the critical parameter, for which interfragmentary motion (IFM) is a controlling variable. Whilst fixator frames are quite stiff, their contribution to the overall fixation stiffness is overshadowed by the low stiffness of the crossed wires or half-pins used to connect to the bone segments [4,27-29].

Our tests show that the two hexapod fixators, TLHEX and TSF, demonstrate similar mechanical behavior, especially at the crossover from tensile to compressive loading. This transition typically occurs during the gait cycle, even if the fractured limb is only being toe-touch loaded. Indeed, our tests show that, due to the hexapod geometry and design details of the struts and their attachment, the mechanical behavior of the assembled hexapod is non-linear (Figure 4), only becoming approximately linear at higher loadings. At higher loadings, we found TSF stiffness of the order of 1700 N.mm^-1, which correspond to values reported by Nikonovas with his “stiff” set up [4], but are much stiffer than those reported by Fenton and colleagues for four ring frames [24]. Indeed, this is sometimes quoted as the advantage of circular fixator systems [23]. The TSF demonstrates 0.8 mm slack or backlash between the linear stiffness regions, which range from 300 Nmm^-1 initially to 923 Nmm^-1 at maximum (Table 2). The TLHEX shows 0.6 mm motion between stiffnesses of 215 Nmm^-1 initially and 1552 Nmm-1 max. This means that describing the system by a single stiffness is not meaningful or useful [4].

When combined with bone analogues (GFRP tubes) and crossed wire fixation, the hexapod/bone construct stiffnesses are again almost linear, but an order of magnitude lower than the hexapods alone. In passing, it should be noted that the use of bone analogues an order of magnitude less stiff than bone (such as nylon or PE) must also contribute to reduced stiffness of the overall construct. In our tests, the TLHEX construct presents a stiffness of 25-26 Nmm^-1 and backlash of 0.65 mm, whilst the TSF construct demonstrates stiffness of 27-29 Nmm^-1 and backlash of 2.1 mm (Table 3). This is primarily due to the low stiffness of the crossed wires used to connect the bone fragments to the hexapod; half-pins are similarly less stiff than the hexapods themselves [27,29]. So, the stiffness of bone-fixator constructs is generally lower than the frames at the core of the fixation system. Wires and half-pins have been studied frequently [20,28,30]; with some studies reporting similar stiffnesses for wires and halfpins [20,28] whereas other reports suggest differences in axial, bending and torsional loading [29].

Mechanically, external fixators can be represented as an assembly of springs, with each element represented by one spring. Thus, the basic construct for external fixation can be modelled by five springs in series (Figure 6). Although in this representation the bone ends are the extreme ends of the chain, in fact it is the relative motion between the bone fragments which is critical (and in reality these are adjacent to each other). For springs in series (such as these) the total stiffness of the construct (Kt) is defined by Equation 1; added elements contribute in a non-linear fashion.

Bone Contacting Elements (wires and half-pins) at one fragment, however, act like springs in parallel and the stiffnesses are simply additive (Equation 2). For them, adding more elements simply adds to the total stiffness. Thus, in our tests, adding a second pair of crossed wires doubles the stiffness of the construct, and adding a further pair of half-pins triples the stiffness.

However, as equation 1 shows, the effect on the total construct is more complex, since the hexapod is in series with the Bone Contacting Elements (BCE). As Equation 1 shows, the combination of all the various components can be summed by the inverse of the stiffness, called the Compliance. In that case, the components in series then deliver an overall. Compliance that is the sum of the individual element compliances: so, we propose that a more useful description of Fixator mechanics is Compliance. Since the effective mechanics of the whole fixator construct are more strongly influenced by the least stiff elements (the bone-to-frame connections- Half Pins and Tensioned wires) than by the hexapod stiffness, small differences between the hexapod behavior do not significantly affect the conditions of fracture healing. The large compliances of the Bone Contacting Elements overpower the small compliance of the hexapod.

In fact, for fracture healing the parameter of most interest is the movement permitted at the fracture site (the Inter- Fragmentary Motion or IFM), which bears an inverse relationship to the fixator stiffness and a direct relationship to the applied load. Furthermore, it is clear from our findings and those of others [4,31-33], that loading wires above their elastic limit decreases their effective stiffness [34], and so limiting weightbearing in crossed-wire constructs is also important.

We postulate that a more meaningful description of the fixator mechanics would thus be to report 15%Compliance or 30%Compliance, where 15%Compliance represents “Toe Touch Weight bearing” [35] or 100N axial loading, and 30%Compliance represents “Partial Weightbearing” or 200N axial loading. Indeed, this measure has already been previously reported [22]. On this basis, the testing in this study indicates that the TLHEX with single rings and crossed 1.8 mm wires provides Toe Touch Compliance (100N: TTC) of 1.45 mm and Partial Weightbearing (PWB) of 1.8 mm. The TSF system similarly set up delivers Toe Touch Compliance (100N) of 1.25 mm and Partial Weightbearing of 1.7 mm. In our testing, a construct with two rings and two pairs of wires on each side shows double the stiffness, so the Compliance is halved. Adding further half-pins triples the stiffness, further reducing these compliances.

It should also be noted that effective fracture healing requires optimal axial strain in the fracture zone, but also minimal shear due to bending or torsional loading. Hexapods are much stiffer in torsion or bending than Ilizarov frames and also planar systems (except in their major plane) [22]. Again, the contribution of the bone-to-frame connections to shear or torsional compliance overrides that of the hexapods, so that any difference between hexapods is insignificant.

Conclusion

1.1. Both hexapods (TSF and TLHEX) demonstrate similar mechanical behavior under physiological loadings.
1.2. External fixator dynamics could usefully be reported by Compliance, which represents the actual bone fragment movement permitted by the fixator: Toe Touch Compliance (at 100N) represents one such measure: in this study the TSF construct gave 1.25 mm whilst the TLHEX gave 1.45 mm in this measure.
1.3. The compliance of the bone-to-frame connections in axial, shear or torsional loading overrides that of the hexapods, so that any difference between hexapods is insignificant in the final mechanics of fixation constructs. Surgeons balance all of these factors, plus the medical complications, to achieve optimal conditions for fracture healing. Bioengineers have a responsibility to refine our understanding and provision of optimal mechanics of all fixation components to support this.

Acknowledgement

We thank Smith & Nephew Richards, Memphis, USA (TSF) and Orthofix S.R.L., Bussolengo, Verona, Italy (TLHex) for the loan of equipment.

Competing interests

None declared.

Funding

None.

Ethical approval

Not required.

All of the work described here was performed at the Department of Bioengineering, Imperial College London.

References

1. Briggs BT, Chao EYS (1982) The Mechanical Performance of the Standard Hoffmann-Vidal External Fixation Apparatus. J Bone Joint Surg Am 64(4): 566-573.
2. Fleming B, Paley D, Kristiansen T, Pope M (1989) A Biomechanical Analysis of the Ilizarov External Fixator. Clin Orthop Relat R 241: 95-105.
3. Kummer F (1989) Biomechanics of the Ilizarov external fixator. Bull Hosp Jt Dis Orthop Inst 383(49): 140-147.
4. Nikonovas A (2005) Taylor Spatial Frame: Kinematics, Mechanical Properties and Automation. Bristol, England.
5. Chao EYS, Aro HT, Lewallen DG, Kelly PJ (1989) The Effect of Rigidity on Fracture-Healing in External Fixation. Clin Orthop Relat R 241: 24-35.
6. Kenwright J, Goodship AE (1989) Controlled mechanical stimulation in the treatment of tibial fractures. Clin Orthop Relat Res (241): 36-47.
7. Wu JJ, Shyr HS, Chao EYS, Kelly PJ (1984) Comparison of Osteotomy Healing under External Fixation Devices with Different Stiffness Characteristics. J Bone Joint Surg Am 66A(8): 1258-1264.
8. Goodship AE, Cunningham JL, Kenwright J (1998) Strain rate and timing of stimulation in mechanical modulation of fracture healing. Clin Orthop Relat Res (355 Suppl): S105-115.
9. Goodship AE, Kenwright J (1985) The influence of induced micromovement upon the healing of experimental tibial fractures. J Bone Joint Surg Br 67(4): 650-655.
10. Kenwright J, Richardson JB, Goodship AE, Evans M, Kelly DJ, et al. (1986) Effect of controlled axial micromovement on healing of tibial fractures. Lancet 2(8517): 1185-1187.
11. Orbay G, Frankel V, Kummer F (1992) The effect of wire configuration on the stability of the Ilizarov external fixator. . Clin Orthop Relat Res 279: 299-302.
12. Gao XS, Lei DL, Liao QZ, Zhang GF (2005) Generalized Stewart-Gough platforms and their direct kinematics. Ieee T Robot 21(2): 141-51.
13. Stewart D (1965) A Platform with Six Degrees of Freedom. Proc Instn Mech Engnrs 180(15): 371 401-486.
14. Al-Sayyad M (2006) Taylor Spatial Frame in the treatment of pediatric and adolescent tibial shaft fractures. J Pediatr Orthoped 26(2): 164-170.
15. Feldman D, Madan S, Koval K, van Bosse H, Bazzi J, et al. (2003) Correction of tibia vara with six-axis deformity analysis and the Taylor Spatial Frame. J Pediatr Orthopedics 23(3): 387-391.
16. Rozbruch S, Helfet D, Blyakher A (2002) Distraction of hypertrophic nonunion of tibia with deformity using Ilizarov/Taylor Spatial Frame. Report of two cases. Arch Orthop Trauma Surg 122(5): 295-298.
17. Sluga M, Pfeiffer M, Kotz R, Nehrer S (2003) Lower limb deformities in children: two-stage correction using the Taylor spatial frame. J Pediatr Orthop B 12(2): 123-128.
18. Gardner T, Evans M (1992) Relative Stiffness, Transverse Displacement and Dynamization in Comparable External Fixators. Clin Biomech 7: 231-239.
19. Gardner T, Evans M, Simpson A, Turner-Smith A (1994) Three-dimensional movement at externally fixated tibial fractures and osteotomies during normal patient function. Clin Biomech (Bristol, Avon) 9: 51-59.
20. Gessmann J, Citak M, Jettkant B, Schildhauer T, Seybold D (2011) The influence of a weight-bearing platform on the mechanical behavior of two Ilizarov ring fixators: tensioned wires vs. half-pins. J Orthopaedic Surg Res 6: 61-71.
21. Smitham P, Khan W, Hazlerigg A, Bajaj S, McCarthy I, et al. (2012) Defining the rattle: A mechanical study of three different types of limb reconstruction frames. J Bone Joint Surg 94B.
22. Henderson D, Rushbrook J, Harwood P, Stewart T (2017) What Are the Biomechanical Properties of the Taylor Spatial Frame (TM). Clin Orthop Relat R 475: 1472-1482.
23. Yilmaz E, Belhan O, Karakurt L, Arslan N, Serin E (2003) Mechanical performance of hybrid Ilizarov external fixator in comparison with Ilizarov circular external fixator. . Clin Biomech 18(6): 518-522.
24. Fenton C, Henderson D, Samchukov M, Cherkashin A, Sharma H (2021) Comparative Stiffness Characteristics of Ilizarov- and Hexapod-type External Frame Constructs. Strateg Trauma Limb Reconstruc 16(3): 138-143.
25. Baran O, Havitcioglu H, Tatari H, Cecen B (2008) The stiffness characteristics of hybrid Ilizarov fixators. J Biomech 41(14): 2960-2963.
26. Hansen U, Zioupos P, Simpson R, Currey JD, Hynd D (2008) The effect of strain rate on the mechanical properties of human cortical bone. J Biomech Eng 130(1): 011011.
27. Antoci V, Voor M, Antoci V, Roberts C (2007) Effect of wire tension on stiffness of tensioned fine wires in external fixation: a mechanical study. Am J Orthop (Belle Mead NJ) 36: 473-476.
28. Khurana A, Byrne C, Evans S, Tanaka H, Haraharan K (2010) Comparison of transverse wires and half pins in Taylor Spatial Frame: a biomechanical study. J Orthopaedic Surg 5: 23.
29. Roberts C, Antoci V, Antoci V, Voor M (2005) The effect of transfixion wire crossing angle on the stiffness of fine wire external fixation: a biomechanical study. Injury 36: 1107-1112.
30. Zamani A, Oyadiji S (2010) Theoretical and Finite Element Modeling of Fine Kirschner Wires in Ilizarov External Fixator. . J Med Devices 4: 1-9.
31. Aronson J, Harp JH (1992) Mechanical considerations in using tensioned wires in a transosseous external fixation system. . Clin Orthop Relat Res 280: 23-29.
32. Delprete C, Gola M (1993) Mechanical performance of external fixators with wires for the treatment of bone fractures--Part II: Wire tension and slippage. J Biomech Engineer 115: 37-42.
33. Hillard P, Harrison A, Atkins R (1998) The yielding of tensioned fine wires in the Ilizarov frame. Proc 374 Inst Mech Eng H 212: 37-47.
34. Watson M, Matthias K, Maffulli N, Hukins D (2003) Yielding of the clamped-wire system in the Ilizarov external fixator. Proc Inst Mech Eng H 217(5): 367-374.
35. Ruedi T, Buckley R, Moran C (2007) AO Principles of Fracture Management. New York, USA. Thieme.