- Email: [email protected]
Biomechanical analysis of distal femoral fracture fixation: dynamic condylar screw versus locked compression plate
J Orthop Sci DOI 10.1007/s00776-014-0583-6
ORIGINAL ARTICLE
Biomechanical analysis of distal femoral fracture fixation: dynamic condylar screw versus locked compression plate Nidhi Narsaria • Ashutosh K. Singh Amit Rastogi • Vakil Singh
•
Received: 2 October 2013 / Accepted: 1 May 2014 The Japanese Orthopaedic Association 2014
Abstract Background This human cadaveric study introduces a laboratory model to establish and compare the fixation stability of the distal femoral locking plate (DFLP) and dynamic condylar screw (DCS) in distal femoral fracture fixation. Materials and methods The study was conducted on 16 fresh cadaveric femoral specimens, 8 implanted with the DCS and the other 8 with the DFLP. The construct was made unstable by removing a standard-sized medial wedge with a 1-cm base (gap osteotomy) beginning 6 cm proximal to the lateral joint line in the distal metaphyseal region with loss of the medial buttress. Each specimen underwent axial and torsional stiffness testing along with cyclic axial loading to failure. The mean DEXA value for the DFLP group was 0.82 g/cm2 and in the DCS group was 0.79 g/cm2. Results Axial stiffness in the DFLP group was significantly higher than in the DCS group, but no significant difference was found in torsional stiffness between the groups. A significant difference was found in the load-tofailure results between the groups. Plastic and total deformation was significantly higher in constructs in the DCS
N. Narsaria A. K. Singh (&) Mayo Institute of Medical Sciences, Barabanki 226010, UP, India e-mail: [email protected] A. Rastogi Department of Orthopedics, Institute of Medical Sciences, BHU, Varanasi, UP, India V. Singh Department of Metallurgy, Institute of Technology, BHU, Varanasi, UP, India
group than in those in the DFLP group. Total energy absorbed before construct failure was also significantly higher in the DFLP group than in the DCS group. Conclusions The DFLP construct proved stronger than the DCS in both axial stiffness and cyclic loading, but similar in torsional stiffness in biomechanical testing in a simulated A3 distal femoral fracture.
Introduction Distal femoral fractures are relatively uncommon and complex fractures accounting for about 7 % of all femoral fractures [1, 2]. The incidence is highest in females older than 75 years and in adolescent males and males 15–24 years old [2]. These fractures are often unstable, comminuted and/or intra-articular, and they usually occur in elderly or multiply-injured patients. These factors place high demands on any surgical implant used to fix these fractures and may lead to failure. The Muller AO/OTA classification is the preferred system [3] and of immediate relevance in guiding the appropriate selection of surgical approaches and implants for specific injuries. For a long time, the angle blade plate (ABP) and compression screw and side plate devices, such as the dynamic condylar screw (DCS), were the most commonly used implants to fix these fractures. Recent trends in fracture care with an emphasis on ‘‘biologic fixation’’ have led to the development of locked compression plates and the less invasive stabilization system (LISS; Synthes), which were developed to minimize fracture site soft-tissue dissection while maximizing fixation and stability of fractures of the distal femur [4, 5]. Other available options for internal fixation of these fractures have many drawbacks. Insertion of blade plates is technically demanding; DCS and ABP require removal of a
123
N. Narsaria et al.
large amount of bone for insertion; condylar buttress plates lack the stability of fixed angle devices and are prone to varus collapse or screw failure [6, 7]. Retrograde intramedullary nails do not sufficiently stabilize fragmented articular fractures [8, 9]. Few biomechanical studies comparing the relative strength of DCS and DFLP fixation in distal femoral fractures are reported in the literature, and they report controversial results. Some studies have shown that the DCS provides stiffer and stronger fixation [10], while others have shown that LCP fixation is stronger than DCS fixation in distal femoral fractures [11, 12]. Hereby, we performed a cadaveric study to evaluate the relative strength and stability of two different implants (DCS and DFLP constructs) by measuring the propensity to biomechanical failure of each design type in vitro. In this study, we aim to establish modes of failure for each device tested and to correlate these with commonly seen fracture patterns clinically.
Materials and methods This research study was done at Institute of Medical Sciences and Department of Metallurgical Engineering, Banaras Hindu University (BHU), India, and approved by the ethics committee and institutional review board of BHU before commencement of the study. This was a cadaveric study so there was no need for informed consent. This study was performed in accordance with the ethical standards of the 1964 Declaration of Helsinki as revised in 2000. Eight pairs of freshly harvested human cadaveric femora were selected. A DCS and DFLP were implanted in eight specimens each; all fixation was done under image intensification. For both the plate material used and screw design (Yogeshwar Private Ltd., Mumbai, India), the implants used in both groups were made by the same manufacturer for consistency. All implants used were made of 3l6L stainless steel. Eight femora were implanted with DCSs using the AO principles of internal fixation. Condylar screws were placed in the distal fragment, and four standard bicortical 4.5-mm screws were placed in the proximal fragment. Another eight femora were implanted with DFLPs. The locking plate, which was first provisionally applied to the lateral surface of the distal fragment, dictated the screw position. These specimens were fixed with five locking screws in the distal fragment and four locking screws in the proximal fragment through the plate. The construct was made unstable by removing a standard-sized medial wedge with a 1-cm base (gap osteotomy) with the help of a standard cutting zig beginning 6 cm proximal to the lateral joint line in the distal metaphyseal region. This established model is meant to simulate an OTA/AO type A3 distal femoral fracture with loss of the medial buttress (Fig. 1a–d).
123
Each specimen underwent axial and torsional stiffness testing along with the cyclic axial loading to failure test using a completely computer-controlled servo-hydraulic MTS testing machine (model 810) of ±50 kN capacity. The testing apparatus for the axial stiffness and cyclic axial loading to failure tests is shown in Fig. 2. The load was applied to the femur through a custom mold. The condyles of the distal end of the femur were also held in a custom mold attached to the material testing machine. In this testing, the line of action of the axial force simulated physiologic loading in the single-leg stance during walking. Axial preload of 100 N was applied proximally to stabilize the construct. Then constructs were loaded in compression at a loading rate of 10 mm/min. Axial loading was performed in a displacement control mode. Testing was stopped when 500 N was reached. Then, torsional stiffness testing was performed. The proximal end was held in a custom mold, and the distal end was secured in a chuck (Fig. 3). Precise positioning was done to ensure that the femoral axis was aligned with the axis of rotation. The custom proximal fixture was mounted to a bearing system. The MTS imparted a force to a lever attached to the bearing to allow rotation of the femur about its longitudinal axis. Torsional loading was performed by preloading each specimen to 5 Nm of torque and then loading the specimen to a maximum torque of 20 Nm at a rate of 25 degrees/min. Cyclic axial loading to failure was performed last with a preload of 100 N, and each specimen underwent 10 cycles at each peak load from 300 to 1000 N in 100-N increments at a rate of 0.75 mm/s as previously described for the distal femur [13, 14]. Each load cycle was conducted as a ramp load in displacement control at a rate of 0.75 mm/s to the peak load of the cycle. The construct was considered to have failed if the implant or femur model fractured, if the implant pulled out of the femur, if the medial edges of the osteotomy closed, or if there was irreversible deformation present on completion of the cyclic axial loading protocol. Residual gap at the fracture site was measured by a direct measuring device, vernier calipers, after cyclic loading. Axial stiffness and torsional stiffness were calculated from the load–displacement curve and torque–angle graph, respectively. For cyclic axial loading to failure, a time– displacement curve was plotted [15]. Reversible and irreversible deformation in cyclical axial loading was calculated for each construct from the time-displacement curve. Plastic (irreversible) deformation was calculated by subtracting the amount of displacement present at the start of the first cycle (300 N) from displacement present after the final cycle. Total deformation was recorded after the last testing cycle [15]. All the statistical analyses were performed using InStat software for Windows (GraphPad version 3.00, San Diego, CA, USA). Student’s t test was used to analyze the mean
DCS versus DFLP fixation in distal femoral fractures Fig. 1 a Bone implant construct showing the cadaveric femur instrumented with the distal femur locking plate (DFLP); metaphyseal osteotomy was done. b Lateral view of the DFLP-bone construct. c Bone implant construct showing the cadaveric femur instrumented with the dynamic condylar screw (DCS); metaphyseal osteotomy was done. d Lateral view of the DCS-bone construct
difference of bone mineral density, axial stiffness, torsional stiffness and cyclic load-to-failure variables (reversible and irreversible deformation, total energy absorbed before construct failure). The test was referenced for the twotailed p value, and the 95 % confidence interval was constructed around the sensitivity proportion using the normal approximation method. A value of \0.05 was considered statistically significant.
Results The mean BMD value of cadaveric femora in the DFLP group was 0.82 g/cm2 and in the DCS group was 0.79 g/ cm2. There was no significant difference between the two
groups (p = 0.502). No visual loss of fixation or complete closure of the medial fracture gap occurred in either the axial or the torsional loading groups. The mean axial stiffness in the DFLP group (72.8 ± 8.6 N/m) was 26.0 % higher than in the DCS group (53.8 ± 4.2 N/m). This was a statistically significant difference (p = 0.002). An insignificant difference (p = 0.286) was found between the torsional stiffness of DFLP constructs (1.78 ± 0.05 Nm/ degree) and DCS constructs (1.62 ± 0.03 Nm/degree) (Table 1). DFLP group constructs (0.8 ± 0.2 mm) had 56 % less irreversible deformation than the DCS group constructs (1.8 ± 0.4 mm), and this difference was statistically significant (p = 0.003). Total deformation in the DFLP group was 12.8 ± 3.2 mm (range 9.4–16.8 mm) and in the DCS group was 16.4 ± 4.8 mm (range
123
N. Narsaria et al. Table 1 Comparison of stiffness between groups Parameters
DFLP group
DCS group
p value
Bone mineral density (g/cm2) Mean
0.942
0.824
SD
0.34
0.28
0.502
Axial stiffness (N/mm) Mean SD
72.8
53.8
8.6
4.2
0.002
Torsional stiffness (Nm/degree) Mean
1.78
1.62
SD
0.05
0.03
0.286
Table 2 Comparison of cyclic load-to-failure results in both groups Load-to-failure variables
DFLP group
DCS group
p value
1.8
0.003
Plastic (irreversible) deformation (mm) Mean
0.8
SD 0.2 0.4 Total (reversible ? irreversible) deformation (mm) Fig. 2 MTS machine with a mounted specimen for axial stiffness and cyclic load-to-failure testing
Mean
12.8
16.4
SD
3.2
4.8
0.002
Energy absorbed before cyclic loading to failure (Newton) Mean
8872
6462
SD
886
640
0.012
osteotomy site, plate barrel deformation of 2–5 and closure of the fracture gap.
Discussion
Fig. 3 MTS machine with a mounted specimen for torsional stiffness testing
11.6–22.6 mm). This difference was also statistically significant (Table 2). During loading to failure, all constructs failed at the osteotomy site. In the DFLP group, all the specimens had irreversible bending of the implant at the osteotomy site and collapse of the fracture site. Three of eight specimens in the DFLP group had implant bending of 10, and the other five specimens had implant bending of more than 15 (15–20). None of the constructs in the DFLP group had screw or implant fracture or screw and implant pullout. Of eight DCS constructs, one had implant and screw pullout along with closure of the medial osteotomy gap. Four DCS constructs had plate bending of 15 at the osteotomy site with closure of the fracture gap. The remaining three constructs had plate bending of 15 at the
123
Distal femoral fractures with metaphyseal bone loss present a challenging problem for orthopedic surgeons [11, 16]. Available implant options for internal fixation of these fractures are cancellous screws, ABPs, condylar buttress plates (CBPs), DCSs, locking condylar plates (LCPs), LISS, retrograde interlock nails and antegrade interlocking nails (IMNs). Previously, the DCS and ABP were the favored implants, but currently locking plates and LISSs are more commonly used [11]. Minimally invasive locking-plate technology has improved the fixation strength of the distal fracture segment because of less bone destruction and more screws being secured on the bent plate [17]. To assess their performance, orthopedic implants are often attached to cadaveric or synthetic bone specimens, and biomechanical testing of the implant-bone construct is performed. Cyclic loading gradually alters the mechanical properties of bone and promotes fatigue fractures; therefore, this is commonly used to investigate the strength of bone-implant constructs. In our study, cyclical testing
DCS versus DFLP fixation in distal femoral fractures
involved axial and torsional loads. Presence of an osteotomy gap was clinically comparable to bone loss in unstable distal femoral fractures; therefore, our loading model primarily tested the stability of the bone-implant construct in AO type 33 A3 fractures. Our study compared the fixation strength of the DFLP and DCS in a cadaveric model of a distal femoral fracture and demonstrated a significant difference in postcycling subsidence between the two constructs. The DFLP showed a mean of 1.0 mm (or 56 %) less subsidence than the DCS after cycling, and this difference was statistically significant. Out of eight DCS specimens, one construct had implant failure in the form of implant pullout after cyclic loading to failure. All other constructs had construct failure in the form of closure of the medial osteotomy gap and implant bending after cyclic loading to failure. Our experiment did not demonstrate a significant difference in the bone density of the specimens in the two groups studied. Further, there was no correlation between bone density and the relative performance of the constructs in load-to-failure testing. It is postulated that this superiority of the DFLP is directly attributed to its biomechanical advantage since multiple screws can be inserted in distal fragments and provide better angular stability. For the distal femur, the angular stability of the distal screws help to prevent varus collapse. The locking screws may also provide stronger fixation of the plate in the proximal fragment by eliminating any potential for toggle and sequential screw loosening. This could have a particular advantage in osteoporotic bones. In addition to this, the locking plate is not compressed against a cortex; therefore, the periosteal blood supply may be preserved. Multiple studies have been done to compare the relative stability of various options and fixed angle plating for supracondylar femoral fractures both mechanically and clinically [12, 18]. The superior fixation strength of locking screws compared to conventional screws has been well established for metaphyseal fracture fixation especially in osteoporotic bones, whereby long unicortical screws act as columns capable of supporting fracture fragments at a fixed angle to the plate [19]. Marti et al. [15] compared the LISS to condylar buttress plate (CBP) and the DCS in distal femoral fracture fixation and showed an enhanced ability to withstand high loads and less irreversible deformation in the LISS than in the DCS and CBP. Bong et al. [20] evaluated the fixation of supracondylar femoral fractures in cadaveric specimens in a 10-mm gap fracture model using a retrograde nail or LISS plate. They found the retrograde nail to be more stable in comminuted fracture patterns, especially in torsion. Higgins et al. [11] in their study compared the strength of the fixed angle blade plate to that of the locking condylar plate and found the latter to be a significantly stronger construct. Stoffel et al. [21, 22] obtained in a bridge plating model of supracondylar
femoral fractures in osteopenic femurs. They reported that locked constructs had a 26 % lower torsional rigidity than non-locked constructs, while the axial stiffness did not significantly differ between the locked and non-locked constructs. Zlowodzki et al. [14] concluded that the fixation strength (load/moment to failure) of the LISS constructs was 34 % greater in axial loading and 32 % less in torsional loading compared with the ABP constructs and 13 % greater in axial loading and 45 % less in torsional loading compared with the IMN constructs. Cyclical axial loading demonstrated significantly less plastic deformation for the LISS construct compared with ABP constructs and similar plastic deformation compared with IMN constructs. Cyclical axial testing demonstrated that the degree of permanent deformation seen after a relatively small number of cycles was significantly higher for the ABP construct compared with the LISS. Comparing the LISS with the retrograde IMN in cyclical axial loading reveals similar permanent deformation. Harder et al. [17] concluded that there was no relevant difference in the mechanical properties of the two fixations (DCS and nonlocked CBP) for fractures without medial defects, even though the stability of the fixation was reduced by removing the distal screw. Furthermore, interfragmental movement was minimal. For simple Y-osteotomies, the CBP did not offer any technical or mechanical advantages. The stability in the frontal plane however was significantly reduced in osteotomies with medial defects. The amplitude of interfragmental movement in all bone constructs was greater than in those fixed by the DCS. A number of other studies have compared the biomechanical properties of these implants in human cadavers with variable results [12, 18, 23, 24]. These studies suggest that locked plates may have some biomechanical advantages over other methods. ABP and DCS constructs seem to behave similarly in extra-articular fractures. The DCS may perform better for some intra-articular fractures. The supracondylar nail (SCN) is not as strong as the DCS or ABP in torsion, but performs similarly in axial loading and anteroposterior bending tests when used for extra-articular fractures. Possible limitations to our experiment lie in the study design. •
•
An inherent limitation of mechanical studies is their inability to accurately reproduce both the internal and external loading environment of the distal femur. There is no accounting for the soft-tissue envelope or bone healing, which is difficult to examine in an in vitro model. Only axial and torsional loading was tested. Media/ lateral and flexion/extension bending of the constructs was not tested. Previous biomechanical studies have
123
N. Narsaria et al.
•
shown torsional loading and medial/lateral bending not to be the mode of failure in the early postoperative period [12, 24]. Another limitation is measuring the BMD of bone by DEXA in relation to the surrounding medium. In retrospect, when measuring the BMD of the specimen, higher values have been obtained by having air as the surrounding medium.
Therefore, these data must be interpreted as strictly biomechanical, representing only part of the scenario at work in the fixation and healing of these injuries in vivo. Although this model did not take into account the actual muscle forces acting in the distal femur, we feel that it was appropriate for comparing the relative stability and stiffness of the two construct groups. When considering micromotion and construct stiffness, the DFLP had statistically significant higher axial stiffness and significantly lower micromotion across the fracture gap with axial loading. However, no significant difference was found in terms of torsional stiffness, but this is clinically irrelevant because patients are often not allowed to perform activities in the postoperative period that lead to torsional loading to the implant. Nonetheless, the significant findings of increased strength of fixation over the DCS certainly appear to support the use of these locking implants clinically. Conflict of interest of interest.
The authors declare that they have no conflict
References 1. Kolmert L, Persson BM, Romanus B. An experimental study of devices for internal fixation of the distal femoral fractures. Clin Orthop. 1982;171:290–9. 2. Martinet O, Cordey J, Harder Y, Maier A, Bu¨hler M, Barraud GE. The epidemiology of fractures of the distal femur. Injury. 2000;31:C62–3. 3. Muller ME, Nazarian S, Koch P, Schatzker J. The comprehensive classification of fractures of long bones. Berlin: Spring; 1990. p. 116–47. 4. Frigg R, Appenzeller A, Christensen R, Frenk A, Gilbert S, Schavan R. The development of the distal femur less invasive stabilization system (LISS). Injury. 2001;32:SC24–31. 5. Krettek C, Muller M, Miclau T. Evolution of minimally invasive plate osteosynthesis (MIPO) in the femur. Injury. 2001;32:SC14– 23. 6. Giles JB, De Lee JC, Heckman JD, Keever JE. Supracondylarintercondylar fractures of the femur treated with a supracondylar plate and lag screw. J Bone Joint Surg Am. 1982;64:864–70. 7. Sanders R, Swiontkowski M, Rosen H, Helfet D. Double-plating of comminuted, unstable fractures of the distal part of the femur. J Bone Joint Surg Am. 1991;73:341–6.
123
8. Firoozbakhsh K, Behzadi K, De Coster TA, Moneim MS, Naraghi FF. Mechanics of retrograde nail versus plate fixation for supracondylar femur fractures. J Orthop Trauma. 1995;9:152–7. 9. Helfet DL, Lorich DG. Retrograde intramedullary nailing of supracondylar femoral fractures. Clin Orthop Relat Res. 1998;350:80–4. 10. Heiney JP, Barnett MD, Vrabec GA, Schoenfeld AJ, Baji A, Njus GO. Distal femoral fixation: a biomechanical comparison of trigen retrograde intramedullary (i.m.) nail, dynamic condylar screw (DCS), and locking compression plate (LCP) condylarplate. J Trauma. 2009;66:443–9. 11. Higgins TF, Pittman G, Hines J, Bachus KN. Biomechanical analysis of distal femur fracture fixation: fixed-angle screw-plate construct versus condylar blade plate. J Orthop Trauma. 2007;21:43–6. 12. Koval KJ, Hoehl JJ, Kummer FJ, Simon JA. Distal femoral fixation: a biomechanical comparison of the standard condylar buttress plate, a locked buttress plate, and the 95-degree blade plate. J Orthop Trauma. 1997;11:521–4. 13. Mahomed N, Harrington I, Kellam J, Maistrelli G, Hearn T, Vroemen J. Biomechanical analysis of the gamma nail and sliding hip screw. Clin Orthop Relat Res. 1994;304:280–8. 14. Zlowodzki M, Williamson S, Cole PA, Zardiackas LD, Kregor PJ. Biomechanical evaluation of the less invasive stabilization system, angled blade plate, and retrograde intramedullary nail for the internal fixation of distal femur fractures. J Orthop Trauma. 2004;18:494–502. 15. Marti A, Fankhauser C, Frenk A, Cordey J, Gasser B. Biomechanical evaluation of the less invasive stabilization system for the internal fixation of distal femur fractures. J Orthop Trauma. 2001;15:482–7. 16. Lucas SE, Seligson D, Henry SL. Intramedullary supracondylar nailing of femoral fractures: a preliminary report of the GSH supracondylar nail. Clin Orthop. 1993;296:200–6. 17. Harder Y, Martinet O, Barraud GE, Cordey J, Regazzoni P. The mechanics of internal fixation of fractures of the distal femur: a comparison of the condylar screw (DCS) with the condylar plate (CP). Injury. 1999;30:31–9. 18. Ito K, Grass R, Zwipp H. Internal fixation of supracondylar femoral fractures: comparative biomechanical performance of the 958 plate and two retrograde nails. J Orthop Trauma. 1998;4:259–66. 19. Kubiak EN, Fulkerson E, Strauss E, Egol KA. The evolution of lockedplates. J Bone Joint Surg Am. 2006;88:189–200. 20. Bong MR, Egol KA, Koval KJ, Kummer FJ, Su ET, Iesaka K. Comparison of the LISS and a retrograde- inserted supracondylar intramedullary nail for fixation of a periprosthetic distal femur fracture proximal to a total knee arthroplasty. J Arthroplasty. 2002;17:876–81. 21. Stoffel K, Dieter U, Stachowiak G, Gachter A, Kuster MS. Biomechanical testing of the LCP—how can stability in locked internal fixators be controlled? Injury. 2003;34(Suppl 2):B11–9. 22. Stoffel K, Lorenz KU, Kuster MS. Biomechanical considerations in plate osteosynthesis: the effect of plate-to-bone compression with and without angular screw stability. J Orthop Trauma. 2007;21:362–8. 23. Kolmert L, Wulff K. Epidemiology and treatment of distal femoral fractures in adults. Acta Orthop Scand. 1982;53:957–62. 24. Jaakkola JI, Lundy DW, Moore T, Jones B, Ganey TM, Hutton WC. Supracondylar femur fracture fixation: mechanical comparison of the 95 degrees condylar side plate and screw versus 95 degrees angled blade plate. Acta Orthop Scand. 2002;73:72–6.
ORIGINAL ARTICLE
Biomechanical analysis of distal femoral fracture fixation: dynamic condylar screw versus locked compression plate Nidhi Narsaria • Ashutosh K. Singh Amit Rastogi • Vakil Singh
•
Received: 2 October 2013 / Accepted: 1 May 2014 The Japanese Orthopaedic Association 2014
Abstract Background This human cadaveric study introduces a laboratory model to establish and compare the fixation stability of the distal femoral locking plate (DFLP) and dynamic condylar screw (DCS) in distal femoral fracture fixation. Materials and methods The study was conducted on 16 fresh cadaveric femoral specimens, 8 implanted with the DCS and the other 8 with the DFLP. The construct was made unstable by removing a standard-sized medial wedge with a 1-cm base (gap osteotomy) beginning 6 cm proximal to the lateral joint line in the distal metaphyseal region with loss of the medial buttress. Each specimen underwent axial and torsional stiffness testing along with cyclic axial loading to failure. The mean DEXA value for the DFLP group was 0.82 g/cm2 and in the DCS group was 0.79 g/cm2. Results Axial stiffness in the DFLP group was significantly higher than in the DCS group, but no significant difference was found in torsional stiffness between the groups. A significant difference was found in the load-tofailure results between the groups. Plastic and total deformation was significantly higher in constructs in the DCS
N. Narsaria A. K. Singh (&) Mayo Institute of Medical Sciences, Barabanki 226010, UP, India e-mail: [email protected] A. Rastogi Department of Orthopedics, Institute of Medical Sciences, BHU, Varanasi, UP, India V. Singh Department of Metallurgy, Institute of Technology, BHU, Varanasi, UP, India
group than in those in the DFLP group. Total energy absorbed before construct failure was also significantly higher in the DFLP group than in the DCS group. Conclusions The DFLP construct proved stronger than the DCS in both axial stiffness and cyclic loading, but similar in torsional stiffness in biomechanical testing in a simulated A3 distal femoral fracture.
Introduction Distal femoral fractures are relatively uncommon and complex fractures accounting for about 7 % of all femoral fractures [1, 2]. The incidence is highest in females older than 75 years and in adolescent males and males 15–24 years old [2]. These fractures are often unstable, comminuted and/or intra-articular, and they usually occur in elderly or multiply-injured patients. These factors place high demands on any surgical implant used to fix these fractures and may lead to failure. The Muller AO/OTA classification is the preferred system [3] and of immediate relevance in guiding the appropriate selection of surgical approaches and implants for specific injuries. For a long time, the angle blade plate (ABP) and compression screw and side plate devices, such as the dynamic condylar screw (DCS), were the most commonly used implants to fix these fractures. Recent trends in fracture care with an emphasis on ‘‘biologic fixation’’ have led to the development of locked compression plates and the less invasive stabilization system (LISS; Synthes), which were developed to minimize fracture site soft-tissue dissection while maximizing fixation and stability of fractures of the distal femur [4, 5]. Other available options for internal fixation of these fractures have many drawbacks. Insertion of blade plates is technically demanding; DCS and ABP require removal of a
123
N. Narsaria et al.
large amount of bone for insertion; condylar buttress plates lack the stability of fixed angle devices and are prone to varus collapse or screw failure [6, 7]. Retrograde intramedullary nails do not sufficiently stabilize fragmented articular fractures [8, 9]. Few biomechanical studies comparing the relative strength of DCS and DFLP fixation in distal femoral fractures are reported in the literature, and they report controversial results. Some studies have shown that the DCS provides stiffer and stronger fixation [10], while others have shown that LCP fixation is stronger than DCS fixation in distal femoral fractures [11, 12]. Hereby, we performed a cadaveric study to evaluate the relative strength and stability of two different implants (DCS and DFLP constructs) by measuring the propensity to biomechanical failure of each design type in vitro. In this study, we aim to establish modes of failure for each device tested and to correlate these with commonly seen fracture patterns clinically.
Materials and methods This research study was done at Institute of Medical Sciences and Department of Metallurgical Engineering, Banaras Hindu University (BHU), India, and approved by the ethics committee and institutional review board of BHU before commencement of the study. This was a cadaveric study so there was no need for informed consent. This study was performed in accordance with the ethical standards of the 1964 Declaration of Helsinki as revised in 2000. Eight pairs of freshly harvested human cadaveric femora were selected. A DCS and DFLP were implanted in eight specimens each; all fixation was done under image intensification. For both the plate material used and screw design (Yogeshwar Private Ltd., Mumbai, India), the implants used in both groups were made by the same manufacturer for consistency. All implants used were made of 3l6L stainless steel. Eight femora were implanted with DCSs using the AO principles of internal fixation. Condylar screws were placed in the distal fragment, and four standard bicortical 4.5-mm screws were placed in the proximal fragment. Another eight femora were implanted with DFLPs. The locking plate, which was first provisionally applied to the lateral surface of the distal fragment, dictated the screw position. These specimens were fixed with five locking screws in the distal fragment and four locking screws in the proximal fragment through the plate. The construct was made unstable by removing a standard-sized medial wedge with a 1-cm base (gap osteotomy) with the help of a standard cutting zig beginning 6 cm proximal to the lateral joint line in the distal metaphyseal region. This established model is meant to simulate an OTA/AO type A3 distal femoral fracture with loss of the medial buttress (Fig. 1a–d).
123
Each specimen underwent axial and torsional stiffness testing along with the cyclic axial loading to failure test using a completely computer-controlled servo-hydraulic MTS testing machine (model 810) of ±50 kN capacity. The testing apparatus for the axial stiffness and cyclic axial loading to failure tests is shown in Fig. 2. The load was applied to the femur through a custom mold. The condyles of the distal end of the femur were also held in a custom mold attached to the material testing machine. In this testing, the line of action of the axial force simulated physiologic loading in the single-leg stance during walking. Axial preload of 100 N was applied proximally to stabilize the construct. Then constructs were loaded in compression at a loading rate of 10 mm/min. Axial loading was performed in a displacement control mode. Testing was stopped when 500 N was reached. Then, torsional stiffness testing was performed. The proximal end was held in a custom mold, and the distal end was secured in a chuck (Fig. 3). Precise positioning was done to ensure that the femoral axis was aligned with the axis of rotation. The custom proximal fixture was mounted to a bearing system. The MTS imparted a force to a lever attached to the bearing to allow rotation of the femur about its longitudinal axis. Torsional loading was performed by preloading each specimen to 5 Nm of torque and then loading the specimen to a maximum torque of 20 Nm at a rate of 25 degrees/min. Cyclic axial loading to failure was performed last with a preload of 100 N, and each specimen underwent 10 cycles at each peak load from 300 to 1000 N in 100-N increments at a rate of 0.75 mm/s as previously described for the distal femur [13, 14]. Each load cycle was conducted as a ramp load in displacement control at a rate of 0.75 mm/s to the peak load of the cycle. The construct was considered to have failed if the implant or femur model fractured, if the implant pulled out of the femur, if the medial edges of the osteotomy closed, or if there was irreversible deformation present on completion of the cyclic axial loading protocol. Residual gap at the fracture site was measured by a direct measuring device, vernier calipers, after cyclic loading. Axial stiffness and torsional stiffness were calculated from the load–displacement curve and torque–angle graph, respectively. For cyclic axial loading to failure, a time– displacement curve was plotted [15]. Reversible and irreversible deformation in cyclical axial loading was calculated for each construct from the time-displacement curve. Plastic (irreversible) deformation was calculated by subtracting the amount of displacement present at the start of the first cycle (300 N) from displacement present after the final cycle. Total deformation was recorded after the last testing cycle [15]. All the statistical analyses were performed using InStat software for Windows (GraphPad version 3.00, San Diego, CA, USA). Student’s t test was used to analyze the mean
DCS versus DFLP fixation in distal femoral fractures Fig. 1 a Bone implant construct showing the cadaveric femur instrumented with the distal femur locking plate (DFLP); metaphyseal osteotomy was done. b Lateral view of the DFLP-bone construct. c Bone implant construct showing the cadaveric femur instrumented with the dynamic condylar screw (DCS); metaphyseal osteotomy was done. d Lateral view of the DCS-bone construct
difference of bone mineral density, axial stiffness, torsional stiffness and cyclic load-to-failure variables (reversible and irreversible deformation, total energy absorbed before construct failure). The test was referenced for the twotailed p value, and the 95 % confidence interval was constructed around the sensitivity proportion using the normal approximation method. A value of \0.05 was considered statistically significant.
Results The mean BMD value of cadaveric femora in the DFLP group was 0.82 g/cm2 and in the DCS group was 0.79 g/ cm2. There was no significant difference between the two
groups (p = 0.502). No visual loss of fixation or complete closure of the medial fracture gap occurred in either the axial or the torsional loading groups. The mean axial stiffness in the DFLP group (72.8 ± 8.6 N/m) was 26.0 % higher than in the DCS group (53.8 ± 4.2 N/m). This was a statistically significant difference (p = 0.002). An insignificant difference (p = 0.286) was found between the torsional stiffness of DFLP constructs (1.78 ± 0.05 Nm/ degree) and DCS constructs (1.62 ± 0.03 Nm/degree) (Table 1). DFLP group constructs (0.8 ± 0.2 mm) had 56 % less irreversible deformation than the DCS group constructs (1.8 ± 0.4 mm), and this difference was statistically significant (p = 0.003). Total deformation in the DFLP group was 12.8 ± 3.2 mm (range 9.4–16.8 mm) and in the DCS group was 16.4 ± 4.8 mm (range
123
N. Narsaria et al. Table 1 Comparison of stiffness between groups Parameters
DFLP group
DCS group
p value
Bone mineral density (g/cm2) Mean
0.942
0.824
SD
0.34
0.28
0.502
Axial stiffness (N/mm) Mean SD
72.8
53.8
8.6
4.2
0.002
Torsional stiffness (Nm/degree) Mean
1.78
1.62
SD
0.05
0.03
0.286
Table 2 Comparison of cyclic load-to-failure results in both groups Load-to-failure variables
DFLP group
DCS group
p value
1.8
0.003
Plastic (irreversible) deformation (mm) Mean
0.8
SD 0.2 0.4 Total (reversible ? irreversible) deformation (mm) Fig. 2 MTS machine with a mounted specimen for axial stiffness and cyclic load-to-failure testing
Mean
12.8
16.4
SD
3.2
4.8
0.002
Energy absorbed before cyclic loading to failure (Newton) Mean
8872
6462
SD
886
640
0.012
osteotomy site, plate barrel deformation of 2–5 and closure of the fracture gap.
Discussion
Fig. 3 MTS machine with a mounted specimen for torsional stiffness testing
11.6–22.6 mm). This difference was also statistically significant (Table 2). During loading to failure, all constructs failed at the osteotomy site. In the DFLP group, all the specimens had irreversible bending of the implant at the osteotomy site and collapse of the fracture site. Three of eight specimens in the DFLP group had implant bending of 10, and the other five specimens had implant bending of more than 15 (15–20). None of the constructs in the DFLP group had screw or implant fracture or screw and implant pullout. Of eight DCS constructs, one had implant and screw pullout along with closure of the medial osteotomy gap. Four DCS constructs had plate bending of 15 at the osteotomy site with closure of the fracture gap. The remaining three constructs had plate bending of 15 at the
123
Distal femoral fractures with metaphyseal bone loss present a challenging problem for orthopedic surgeons [11, 16]. Available implant options for internal fixation of these fractures are cancellous screws, ABPs, condylar buttress plates (CBPs), DCSs, locking condylar plates (LCPs), LISS, retrograde interlock nails and antegrade interlocking nails (IMNs). Previously, the DCS and ABP were the favored implants, but currently locking plates and LISSs are more commonly used [11]. Minimally invasive locking-plate technology has improved the fixation strength of the distal fracture segment because of less bone destruction and more screws being secured on the bent plate [17]. To assess their performance, orthopedic implants are often attached to cadaveric or synthetic bone specimens, and biomechanical testing of the implant-bone construct is performed. Cyclic loading gradually alters the mechanical properties of bone and promotes fatigue fractures; therefore, this is commonly used to investigate the strength of bone-implant constructs. In our study, cyclical testing
DCS versus DFLP fixation in distal femoral fractures
involved axial and torsional loads. Presence of an osteotomy gap was clinically comparable to bone loss in unstable distal femoral fractures; therefore, our loading model primarily tested the stability of the bone-implant construct in AO type 33 A3 fractures. Our study compared the fixation strength of the DFLP and DCS in a cadaveric model of a distal femoral fracture and demonstrated a significant difference in postcycling subsidence between the two constructs. The DFLP showed a mean of 1.0 mm (or 56 %) less subsidence than the DCS after cycling, and this difference was statistically significant. Out of eight DCS specimens, one construct had implant failure in the form of implant pullout after cyclic loading to failure. All other constructs had construct failure in the form of closure of the medial osteotomy gap and implant bending after cyclic loading to failure. Our experiment did not demonstrate a significant difference in the bone density of the specimens in the two groups studied. Further, there was no correlation between bone density and the relative performance of the constructs in load-to-failure testing. It is postulated that this superiority of the DFLP is directly attributed to its biomechanical advantage since multiple screws can be inserted in distal fragments and provide better angular stability. For the distal femur, the angular stability of the distal screws help to prevent varus collapse. The locking screws may also provide stronger fixation of the plate in the proximal fragment by eliminating any potential for toggle and sequential screw loosening. This could have a particular advantage in osteoporotic bones. In addition to this, the locking plate is not compressed against a cortex; therefore, the periosteal blood supply may be preserved. Multiple studies have been done to compare the relative stability of various options and fixed angle plating for supracondylar femoral fractures both mechanically and clinically [12, 18]. The superior fixation strength of locking screws compared to conventional screws has been well established for metaphyseal fracture fixation especially in osteoporotic bones, whereby long unicortical screws act as columns capable of supporting fracture fragments at a fixed angle to the plate [19]. Marti et al. [15] compared the LISS to condylar buttress plate (CBP) and the DCS in distal femoral fracture fixation and showed an enhanced ability to withstand high loads and less irreversible deformation in the LISS than in the DCS and CBP. Bong et al. [20] evaluated the fixation of supracondylar femoral fractures in cadaveric specimens in a 10-mm gap fracture model using a retrograde nail or LISS plate. They found the retrograde nail to be more stable in comminuted fracture patterns, especially in torsion. Higgins et al. [11] in their study compared the strength of the fixed angle blade plate to that of the locking condylar plate and found the latter to be a significantly stronger construct. Stoffel et al. [21, 22] obtained in a bridge plating model of supracondylar
femoral fractures in osteopenic femurs. They reported that locked constructs had a 26 % lower torsional rigidity than non-locked constructs, while the axial stiffness did not significantly differ between the locked and non-locked constructs. Zlowodzki et al. [14] concluded that the fixation strength (load/moment to failure) of the LISS constructs was 34 % greater in axial loading and 32 % less in torsional loading compared with the ABP constructs and 13 % greater in axial loading and 45 % less in torsional loading compared with the IMN constructs. Cyclical axial loading demonstrated significantly less plastic deformation for the LISS construct compared with ABP constructs and similar plastic deformation compared with IMN constructs. Cyclical axial testing demonstrated that the degree of permanent deformation seen after a relatively small number of cycles was significantly higher for the ABP construct compared with the LISS. Comparing the LISS with the retrograde IMN in cyclical axial loading reveals similar permanent deformation. Harder et al. [17] concluded that there was no relevant difference in the mechanical properties of the two fixations (DCS and nonlocked CBP) for fractures without medial defects, even though the stability of the fixation was reduced by removing the distal screw. Furthermore, interfragmental movement was minimal. For simple Y-osteotomies, the CBP did not offer any technical or mechanical advantages. The stability in the frontal plane however was significantly reduced in osteotomies with medial defects. The amplitude of interfragmental movement in all bone constructs was greater than in those fixed by the DCS. A number of other studies have compared the biomechanical properties of these implants in human cadavers with variable results [12, 18, 23, 24]. These studies suggest that locked plates may have some biomechanical advantages over other methods. ABP and DCS constructs seem to behave similarly in extra-articular fractures. The DCS may perform better for some intra-articular fractures. The supracondylar nail (SCN) is not as strong as the DCS or ABP in torsion, but performs similarly in axial loading and anteroposterior bending tests when used for extra-articular fractures. Possible limitations to our experiment lie in the study design. •
•
An inherent limitation of mechanical studies is their inability to accurately reproduce both the internal and external loading environment of the distal femur. There is no accounting for the soft-tissue envelope or bone healing, which is difficult to examine in an in vitro model. Only axial and torsional loading was tested. Media/ lateral and flexion/extension bending of the constructs was not tested. Previous biomechanical studies have
123
N. Narsaria et al.
•
shown torsional loading and medial/lateral bending not to be the mode of failure in the early postoperative period [12, 24]. Another limitation is measuring the BMD of bone by DEXA in relation to the surrounding medium. In retrospect, when measuring the BMD of the specimen, higher values have been obtained by having air as the surrounding medium.
Therefore, these data must be interpreted as strictly biomechanical, representing only part of the scenario at work in the fixation and healing of these injuries in vivo. Although this model did not take into account the actual muscle forces acting in the distal femur, we feel that it was appropriate for comparing the relative stability and stiffness of the two construct groups. When considering micromotion and construct stiffness, the DFLP had statistically significant higher axial stiffness and significantly lower micromotion across the fracture gap with axial loading. However, no significant difference was found in terms of torsional stiffness, but this is clinically irrelevant because patients are often not allowed to perform activities in the postoperative period that lead to torsional loading to the implant. Nonetheless, the significant findings of increased strength of fixation over the DCS certainly appear to support the use of these locking implants clinically. Conflict of interest of interest.
The authors declare that they have no conflict
References 1. Kolmert L, Persson BM, Romanus B. An experimental study of devices for internal fixation of the distal femoral fractures. Clin Orthop. 1982;171:290–9. 2. Martinet O, Cordey J, Harder Y, Maier A, Bu¨hler M, Barraud GE. The epidemiology of fractures of the distal femur. Injury. 2000;31:C62–3. 3. Muller ME, Nazarian S, Koch P, Schatzker J. The comprehensive classification of fractures of long bones. Berlin: Spring; 1990. p. 116–47. 4. Frigg R, Appenzeller A, Christensen R, Frenk A, Gilbert S, Schavan R. The development of the distal femur less invasive stabilization system (LISS). Injury. 2001;32:SC24–31. 5. Krettek C, Muller M, Miclau T. Evolution of minimally invasive plate osteosynthesis (MIPO) in the femur. Injury. 2001;32:SC14– 23. 6. Giles JB, De Lee JC, Heckman JD, Keever JE. Supracondylarintercondylar fractures of the femur treated with a supracondylar plate and lag screw. J Bone Joint Surg Am. 1982;64:864–70. 7. Sanders R, Swiontkowski M, Rosen H, Helfet D. Double-plating of comminuted, unstable fractures of the distal part of the femur. J Bone Joint Surg Am. 1991;73:341–6.
123
8. Firoozbakhsh K, Behzadi K, De Coster TA, Moneim MS, Naraghi FF. Mechanics of retrograde nail versus plate fixation for supracondylar femur fractures. J Orthop Trauma. 1995;9:152–7. 9. Helfet DL, Lorich DG. Retrograde intramedullary nailing of supracondylar femoral fractures. Clin Orthop Relat Res. 1998;350:80–4. 10. Heiney JP, Barnett MD, Vrabec GA, Schoenfeld AJ, Baji A, Njus GO. Distal femoral fixation: a biomechanical comparison of trigen retrograde intramedullary (i.m.) nail, dynamic condylar screw (DCS), and locking compression plate (LCP) condylarplate. J Trauma. 2009;66:443–9. 11. Higgins TF, Pittman G, Hines J, Bachus KN. Biomechanical analysis of distal femur fracture fixation: fixed-angle screw-plate construct versus condylar blade plate. J Orthop Trauma. 2007;21:43–6. 12. Koval KJ, Hoehl JJ, Kummer FJ, Simon JA. Distal femoral fixation: a biomechanical comparison of the standard condylar buttress plate, a locked buttress plate, and the 95-degree blade plate. J Orthop Trauma. 1997;11:521–4. 13. Mahomed N, Harrington I, Kellam J, Maistrelli G, Hearn T, Vroemen J. Biomechanical analysis of the gamma nail and sliding hip screw. Clin Orthop Relat Res. 1994;304:280–8. 14. Zlowodzki M, Williamson S, Cole PA, Zardiackas LD, Kregor PJ. Biomechanical evaluation of the less invasive stabilization system, angled blade plate, and retrograde intramedullary nail for the internal fixation of distal femur fractures. J Orthop Trauma. 2004;18:494–502. 15. Marti A, Fankhauser C, Frenk A, Cordey J, Gasser B. Biomechanical evaluation of the less invasive stabilization system for the internal fixation of distal femur fractures. J Orthop Trauma. 2001;15:482–7. 16. Lucas SE, Seligson D, Henry SL. Intramedullary supracondylar nailing of femoral fractures: a preliminary report of the GSH supracondylar nail. Clin Orthop. 1993;296:200–6. 17. Harder Y, Martinet O, Barraud GE, Cordey J, Regazzoni P. The mechanics of internal fixation of fractures of the distal femur: a comparison of the condylar screw (DCS) with the condylar plate (CP). Injury. 1999;30:31–9. 18. Ito K, Grass R, Zwipp H. Internal fixation of supracondylar femoral fractures: comparative biomechanical performance of the 958 plate and two retrograde nails. J Orthop Trauma. 1998;4:259–66. 19. Kubiak EN, Fulkerson E, Strauss E, Egol KA. The evolution of lockedplates. J Bone Joint Surg Am. 2006;88:189–200. 20. Bong MR, Egol KA, Koval KJ, Kummer FJ, Su ET, Iesaka K. Comparison of the LISS and a retrograde- inserted supracondylar intramedullary nail for fixation of a periprosthetic distal femur fracture proximal to a total knee arthroplasty. J Arthroplasty. 2002;17:876–81. 21. Stoffel K, Dieter U, Stachowiak G, Gachter A, Kuster MS. Biomechanical testing of the LCP—how can stability in locked internal fixators be controlled? Injury. 2003;34(Suppl 2):B11–9. 22. Stoffel K, Lorenz KU, Kuster MS. Biomechanical considerations in plate osteosynthesis: the effect of plate-to-bone compression with and without angular screw stability. J Orthop Trauma. 2007;21:362–8. 23. Kolmert L, Wulff K. Epidemiology and treatment of distal femoral fractures in adults. Acta Orthop Scand. 1982;53:957–62. 24. Jaakkola JI, Lundy DW, Moore T, Jones B, Ganey TM, Hutton WC. Supracondylar femur fracture fixation: mechanical comparison of the 95 degrees condylar side plate and screw versus 95 degrees angled blade plate. Acta Orthop Scand. 2002;73:72–6.