Biological osteosynthesis is an approach to fracture management that prioritizes the preservation of periosseous soft tissues and blood supply over precise anatomic reconstruction while satisfying mechanical and morphological imperatives to achieve fracture healing and optimal functional outcomes.1 Locking plates are often used in this approach because these implants do not rely on frictional forces between plate and bone that are required with conventional plates to maintain fracture reduction.2 Application of locking plates causes minimal impedance of cortical vascularity3,4 and conserves the periosteum, extraosseous soft tissues, and extraosseous blood supply.3 Locking plates rely on fixed-angle screw rigidity to maintain alignment of the secured bone segments and inter-fragmentary stability.5 These implants have been used effectively in dogs with promising clinical results.4,6–9
The conversion from conventional to locking plate applications has brought about modifications in the size, function, and relationship between plates and screws.5 With conventional plates, the screws secure the plate firmly to the underlying cortex, generating friction between the underside of the plate and the bone, thereby stabilizing the bone segments. The screwhead assumes a largely compressive role through the axial tensile forces within the screw, and the screw thread is matched to the type of bone (cortical vs cancellous) engaged to prevent pullout.10 Conversely, locking plate stability is provided by the angle-stable screws and does not require plate-to-bone contact. The screws function as supportive struts and are exposed to cantilever bending forces concentrated at the interface between the plate and screw, and there is less dependency on osseous screw pullout resistance with locking systems.11 Maintaining the locked relationship between the plate, screws, and secured bone segments for the duration of bone healing is critical.11 This security is dependent on the design of the locking screw-plate mechanism and the quality of the manufacturing process as well as adherence to the correct surgical technique through insertion of the screw at the predefined angle (typically restricted by a purpose-built drill guide) and the application of appropriate insertional torque.12 The locking mechanism can be attained through several design methods, including interdigitation of multiple threads on the screwhead and the plate hole (ie, thread in thread), press fit of the head of a conventional cortical screw into a slightly smaller receiving socket in the plate combined with interdigitation of the screw shaft threads with the screw hole, and press fit of the conical head of a locking screw into a conical screw hole.12
A commercially available 3.5-mm PLS composed of a series of stainless steel spheres or nodes (termed pearls) connected by malleable cylindrical internodes is widely used in veterinary orthopedic surgery.13–16 This design allows multiplanar contouring of the plates, a characteristic that is advantageous when stabilizing bones with irregular topography (eg, for vertebral column stabilization or treatment of fractures of the distal aspect of the femur, the humerus, the acetabulum, or the ilium).13–16 A salient advantage of this PLS system is that, unlike other locking plate systems, the plates accept conventional cortical screws, and the locking mechanism is accomplished by interdigitation of the screw threads with threads in the dependent recess of the screw hole in the nodes as well as a press fit of the screwhead into the conical receptacle in the node.17 A new PLS system is available that uses a similar node and internode conformation but differs from the aforementioned PLS system because it uses screws with a larger core diameter (2.9 mm, compared with a core diameter of 2.4 mm for a conventional 3.5-mm cortical screw) and a thread-in-thread locking mechanism. The latter system affords 3-D plate contouring similar to that accomplished with the previously described construct but theoretically should provide increased bone-to-plate construct stiffness when used clinically (owing to larger core diameter and secure locking mechanism) as well as a decreased risk of implant failure at the level of the screw-plate interface.
Comparing the mechanical properties of the widely used PLS and the newer PLS that incorporates the larger core diameter, locking-head screws and determining the extent to which contouring affects implant stability could aid clinicians in appropriate implant selection and might possibly decrease the risk of implant failure. The purpose of the study reported here was to evaluate and compare mechanical properties (axial stiffness, yield load, failure load and deformation at failure) of these 2 PLS constructs in a synthetic bone model with a simulated fracture gap and to compare screw push-out forces for both PLSs before and after standardized plate contouring. Our first hypothesis was that the PLS constructs that include larger core diameter, locking-head screws would have greater stiffness, yield load, and failure load and less deformation during mechanical testing than the PLS constructs that include conventional cortical screws. We also hypothesized that the uncontoured specimens of the former PLS type would have greater resistance to screw push-out than the latter and that plate contouring would significantly diminish resistance to screw push-out in both PLSs.
Materials and Methods
Construct preparation and axial compression testing
Two PLSs (PLS 1a and PLS 2b) were evaluated in the study. Eight 3.5-mm plate-synthetic bone model constructs were prepared with each type of PLS by 1 investigator (GT). All plates for axial compression testing were uncontoured. Each 8-hole plate segment was applied centrally over a 16.5 × 3-cm-diameter synthetic bonec cylinder. Three bicortical, 3.5-mm screws (conventional cortical screws for the PLS 1 constructs and locking-head screws for the PLS 2 constructs) were inserted through screw holes 1 through 3 in the proximal segment and 6 through 8 in the distal segment, leaving holes 4 and 5 unused. The equipment used was provided by the PLS manufacturers.
To mimic the clinical application technique, all screws were maximally hand-tightened without the use of a torque limiting screwdriver. A 34-mm gap was created18 with an oscillating saw after application of the plate to the synthetic bone model. The gap was centered beneath the 2 central open nodes of each plate. The segmental gap simulated a nonrecon-structed, comminuted, mid-diaphyseal fracture. One end of each construct was then secured in a pot of styrene-acrylic polymer,d with a leveling instrument used to ensure that the long axis of the construct was perpendicular to the base of the pot. The constructs were then fixed to a custom-made jig and secured to the loading cell of a servohydraulic materials testing machinee with clamps (Figure 1). The load cell was adjusted to zero prior to testing, and a preload of 20 N was applied to each construct. The constructs were then tested with a single cycle to failure in axial compression with displacement applied at 1 mm/min. Failure was defined as a 20-mm gap subsidence or catastrophic construct failure.
Photograph of the testing setup for axial compression testing used in an investigation of mechanical properties (stiffness, yield load, failure load, and deformation at failure) of 2 PLS constructs (PLS 1 and PLS 2) in a simulated fracture gap model (8 constructs/group). One of the PLS 1 constructs is shown secured to the servohydraulic materials testing machine.
Citation: American Journal of Veterinary Research 80, 6; 10.2460/ajvr.80.6.533
Axial stiffness, yield load, failure load, and deformation (plate bending) at failure were recorded for each construct. Axial stiffness of the construct was measured from the linear portion of the load-displacement curve between 500 and 700 N for all constructs. This loading range was chosen as these measurements for all constructs had linear behavior within this region, and the range was deemed to be outside of the toe region for all samples. Yield load was determined as the point at which the load-displacement curve deviated from the linear region, as determined by use of the standard 0.2% slope offset criterion. Digital photographs of the plates were obtained after testing when the implants were removed from the synthetic bone model; the degree of bending that occurred when the plate-synthetic bone constructs were tested was quantified by analyzing lateral images of each construct with open-access softwaref to measure the acute angle formed by the intersection of a line drawn parallel to the proximal plate segment and a line drawn parallel to the distal plate segment.
Specimen preparation and screw push-out testing
Screw push-out was evaluated for 12 uncontoured 3-hole plates (6 PLS 1 and 6 PLS 2) and 12 contoured 3-hole plates (6 PLS 1 and 6 PLS 2). A modification of a previously reported method was used to evaluate push-out strength.12 A 3-hole unmodified plate segment was firmly secured to the loading cell of the testing machine with a custom-made jig. A 32-mm-long, 3.5-mm conventional cortical screw (for the PLS 1) or a 32-mm-long, 3.5-mm locking-head screw (for the PLS 2) was secured in the central hole of the plate with the tip of the screw directed toward the actuator of the materials testing machine. The screws were tightened and locked into the plate node to a torque of 1.5 N·m by use of a torque-limiting screwdriver.g A 65-mm-long stainless steel cylinder with a hollow inner diameter of 4 mm was secured to the end of the actuator of the materials testing machine and slid over the protruding screw until 4 mm of the screw below the plate was exposed (Figure 2). A push-out force was applied to the screw parallel to its long axis and perpendicular to the plate at a constant displacement rate of 1 mm/min. The axial load required to unlock the mechanism, defined as an abrupt change in the load-displacement curve, was recorded.
Photograph of the testing setup for screw push-out force of PLS 1 and PLS 2 specimens. After the screw was locked in the central hole of each 3-hole plate at a torque of 1.5 N·m, a load parallel to the screw and perpendicular to the plate (depicted by the arrow) was applied until failure of the locking mechanism was noted on the load-displacement curve. Twelve specimens (6 uncontoured and 6 contoured through a standard approach to achieve a bend of 12.0 ± 0.3°) were tested for each PLS type; the image depicts an uncontoured PLS 1 specimen.
Citation: American Journal of Veterinary Research 80, 6; 10.2460/ajvr.80.6.533
To test the effects of plate contouring on screw push-out, a standard uniplanar mediolateral bend was created in each plate by use of the bending irons supplied with the PLS 1 after insertion and tightening of the nodal tees (as recommended by the manufacturer of that PLS). To achieve uniform bending, each internode was separately bent until the free ends of the bending irons touched. Digital photographs were obtained, and open-source image processing softwaref was used to measure the degree of plate bending. The values were recorded to verify that all specimens underwent the same amount of bending with a tolerance of ± 3° relative to the unmodified plates. The amount of angular deformation achieved after standardized bending of the plates for the push-out testing experiment was 12.0 ± 0.3° for PLS 1 and 12.0 ± 0.3° for PLS 2 plates. A different custom-made jig was used, and the screws were subjected to the same force as for the uncontoured plates. The testing of all specimens was video-recorded to allow posttest evaluation of the mechanism of failure.
Statistical analysis
All statistical analyses were performed with commercially available software.h Normality of data distribution was visually assessed and confirmed on Q-Q plots. Axial stiffness, yield load, failure load, and degree of plate deformation (bending) after axial load testing were expressed as mean ± SD and were compared between the PLS 1 and PLS 2 groups by means of a Student t test. Two-way ANOVA was used to compare the screw push-out force (axial load) between uncontoured and contoured specimens within and between the PLS 1 and PLS 2 groups. Values of P < 0.05 were considered significant.
Results
Axial compression testing
Yield load could only be calculated for 5 of 8 PLS 1 and 4 of 8 PLS 2 constructs. In the remaining constructs, multiple subultimate-failure yielding events with minimal changes in construct stiffness following the event prevented determination of clinically relevant yielding. Mean axial stiffness did not differ between PLS 1 and PLS 2 constructs (P = 0.76; Table 1). Mean yield load and maximum failure load for the PLS 1 constructs were 44.3% and 21% higher than those for PLS 2 constructs, respectively (P < 0.05 for both comparisons).
Comparison of mean ± SD results of mechanical and screw push-out testing for 2 types of PLS constructs (PLS 1 and PLS 2) in a simulated fracture gap model.
| Variable | PLS 1 | PLS 2 | P value |
|---|---|---|---|
| Mechanical testing | |||
| Axial stiffness (N/mm) | 360.3 ± 113.4 | 377.3 ± 109.3 | 0.76 |
| Yield load (N) | 1,538.2 ± 209.8 | 1,066 ± 135.6 | 0.006 |
| Failure load (N) | 1,621.7 ± 252.6 | 1,340 ± 222.3 | 0.03 |
| Screw push-out force (N) | |||
| Uncontoured plates | 2,277.1 ± 300.8 | 4,939.5 ± 893.7 | < 0.001 |
| Contoured plates | 2,062.6 ± 268.7 | 4,292.1 ± 537 | < 0.001 |
Mechanical testing was performed on 16 constructs with uncontoured plates (8 each in the PLS 1 and PLS 2 groups); yield load could only be calculated for 5 constructs in the PLS 1 group and 4 in the PLS 2 group. Screw push-out testing was performed on 24 specimens (6 uncontoured plates and 6 contoured plates/group). For screw push-out testing, the maximum load necessary to unlock screws from the plate was recorded.
Values of P < 0.05 were considered significant.
All constructs failed by plastic deformation of plates without implant breakage. Subjectively, marked deformation of the fourth node screw hole was noted in all PLS 2 group plates. Minimal deformation of the fourth node screw hole was noted in 2 of the PLS 1 group plates. No deformation of the screw was noted in either group. Although angulated in relation to each other, the proximal and distal plate sections of each construct remained straight and affixed (3 screws each) to respective sections of synthetic bone. Mean ± SD angular deformation, measured after removal of the plate from the synthetic bone model, of the PLS 1 plates after axial load-to-failure testing was 6.9 ± 1.7°, which was significantly (P < 0.001) less than that for the PLS 2 plates (14.8 ± 1.7°).
Screw push-out
Mean push-out force to unlock screws from the plate was significantly (P < 0.001 for both comparisons) higher in the PLS 2 group than in the PLS 1 group for both uncontoured and contoured plates (Table 1). No within-group differences in mean screw push-out force were detected between contoured and uncontoured plates (P = 0.26 and P = 0.15 for the PLS 1 and PLS 2 groups, respectively).
Failure of the locking mechanism in PLS 1 group specimens in screw push-out experiments was caused by stripping of the first 2 threads of the screw with minimal deformation of the node for both uncontoured and contoured plates (Figure 3). Failure of the locking mechanism in the PLS 2 group specimens was caused by gross plastic deformation of the node followed by stripping of the last 2 threads at the base of the screwhead for both uncontoured and contoured plates (Figure 4).
Representative photographs depicting the mode of failure of the locking mechanism for PLS 1 specimens. A—Notice the plastic deformation of the first 2 threads of the screw body. B—There is minimal deformation of the node.
Citation: American Journal of Veterinary Research 80, 6; 10.2460/ajvr.80.6.533
Representative photographs depicting the mode of failure of the locking mechanism for PLS 2 specimens. A—Notice the plastic deformation of the last 2 threads of the screwhead. B—There is obvious deformation of the node.
Citation: American Journal of Veterinary Research 80, 6; 10.2460/ajvr.80.6.533
Discussion
The present study revealed no significant difference in stiffness between PLS 1 and PLS 2 constructs during axial loading, but the higher yield load and failure load and lower deformation sustained by the PLS 1 constructs compared with those for PLS 2 constructs, resulted in rejection of our first study hypothesis. According to the manufacturer's specifications, both PLS 1 and PLS 2 implants are made of grade 316L stainless steel and they have the same node and internode dimensions. One possible explanation for the observed biomechanical differences could be a difference in the interaction between the screws and the plates of the 2 systems during axial loading. The screw holes in the PLS 2 plates are 1 mm larger than those in the PLS 1 plates to accommodate the larger core diameter screws. Because the constructs failed by bending, the thinner rim of the nodes in the PLS 2 plates may have been responsible for a decrease in the area moment of inertia at the level of the screw hole, leading to the observed lower yield and failure loads. This speculation was supported by our observation that marked plastic deformation of the fourth node screw hole was noted in all PLS 2 constructs tested, whereas node deformation was only noted in 2 of the PLS 1 constructs and was nominal, compared with that noted in the PLS 2 constructs.
The PLS 2 specimens (which include a thread-in-thread locking mechanism) had a significantly higher screw push-out force than the PLS 1 specimens, which supported our second hypothesis. This finding corroborated results reported by Boero Baroncelli et al.12 In that study,12 implants that had a thread-in-thread locking mechanism had a higher screw push-out force than did the PLS that accepts standard cortical screws (ie, PLS 1 in the present study), and the authors hypothesized that the higher number of threads engaging the plate were responsible for the stronger locking mechanism. However, a minimum screw push-out force for locking plate-and-screw systems considered acceptable for clinical applications has not yet been established. Cyclic testing, which might be more likely to unlock the screw from the plate, might yield more clinically relevant information. Clinical failure of implants due to screw push-out from the plate has not, to our knowledge, been reported for either of the PLSs investigated in the present study. Although a higher screw push-out force may be advantageous, we believe that push-out forces in convalescing small animals are unlikely to exceed the resistance offered by either of these 2 systems.
The method of locking mechanism failure in the PLS 1 group during screw push-out testing in the present study was the same as that previously reported.12 All PLS 1 specimens in our study failed by plastic deformation of the first 2 threads subjacent to the head of the screw with resultant unlocking of the screwhead from the recess of that plate node. The method of failure of the locking mechanism of the PLS 2 specimens was distinct from that of the PLS 1 specimens. Before the screwhead unlocked from the plate by stripping of the last 2 threads at its base, marked plastic deformation of the node was noted during testing and after reviewing the video recordings of the testing. We considered it possible that deformation of the node resulted in uncoupling of the first few threads of the screwhead from the node, inducing failure of the locking mechanism when the remaining 2 engaged threads of the screwhead stripped from the node.
Contouring of the plates did not affect security of the locking mechanism in either system, which caused us to reject our third hypothesis. In a study19 evaluating the effects of plate bending at the level of the screw hole in locking compression plates, a difference was noted in the number of cycles a screw would tolerate prior to loosening, but only if the angular deformation of the plate was > 15°. In our study, plates were deformed 12°, and more pronounced bending may have had an adverse effect on screw security. An additional study evaluating multiple groups of plates with incremental increases in bending angles would be required to substantiate this premise.
We applied a torque of 1.5 N·m when tightening the screws in preparation for screw push-out testing. We were unable to obtain recommended torque figures from the PLS 2 manufacturer. A previous study12 did not find a difference in push-out force for the PLS 1 plates when screws were locked to the plates with 0.8, 1.5, 2.5, and 3.5 N·m of torque. A torque of 1.5 N·m is close to the mean torque value (1.289 N·m) obtained experimentally when 3.5-mm screws are hand tightened.20 The suggested torque when applying 3.5-mm locking screws in 1 popular locking compression plate system is 1.5 N·m, which influenced application practices in our study.
The present study had various limitations. First, the small sample size in each of the testing groups increased the risk of a type II statistical error. We were only able to determine yield loads for 5 of 8 PLS 1 and 4 of 8 PLS 2 constructs because of multiple subultimate-failure yielding events registered during testing of the constructs. Second, constructs were tested in vitro and with only a single load to failure instead of cyclic loading. Although cyclic loading is more representative of in vivo loading conditions,18,21–23 the results of load-to-failure testing provided useful biomechanical information that can be considered when determining clinical applications. Third, we tested the effect of a single mild plate contour in the mediolateral plane only on mechanical performance of the PLSs. Contouring a plate in a single plane helped to maintain consistency during preparation of the specimens; however, surgeons often contour such plates in multiple planes to better conform the implant to the shape of the bone. The effect of multiplanar plate contouring on the strength of the locking mechanisms should be investigated.
Acknowledgments
Funded by a Colorado State University Young Investigator's grant. The PLS 2 implants used in the study were donated by Veterinary Orthopedic Implants (VOI). None of the authors have affiliations, involvement with, or financial interest in the company that provided the PLS 1 or PLS 2 implants. Funding sources did not have any involvement in the study design, data analysis and interpretation, or writing and publication of the manuscript.
ABBREVIATIONS
| PLS | Pearl-type locking plate system |
Footnotes
String-of-Pearls locking plate system, 3.5 mm, OrthoMed, Huddersfield, England.
Locking Cortical Pearl Plate, 3.5 mm, Veterinary Orthopedic Implants, St Augustine, Fla.
Sawbones, Vashon, Wash.
Bondo, 3M, Saint Paul, Minn.
MTS 858 Mini Bionix II, MTS Systems Corp, Eden Prairie, Minn.
ImageJ, version 1.51. National Institutes of Health, Bethesda, Md. Available at: www.imagej.net. Accessed Nov 20, 2017.
Electromatic Equipment Co Inc, Cedarhurst, NY.
Microsoft Excel, version 15.30, Microsoft Inc, Redmond, Wash.
References
1. Palmer RH. Biological osteosynthesis. Vet Clin North Am Small Anim Pract 1999;29:1171–1185.
2. Strauss EJ, Schwarzkopf R, Kummer F, et al. The current status of locked plating: the good, the bad, and the ugly. J Orthop Trauma 2008;22:479–486.
3. Borrelli J Jr, Prickett W, Song E, et al. Extraosseous blood supply of the tibia and the effects of different plating techniques: a human cadaveric study. J Orthop Trauma 2002;16:691–695.
4. Garofolo S, Pozzi A. Effect of plating technique on periosteal vasculature of the radius in dogs: a cadaveric study. Vet Surg 2013;42:255–261.
5. Gautier E, Sommer C. Guidelines for the clinical application of the LCP. Injury 2003;34(suppl 2):B63–B76.
6. Voss K, Kull MA, Hässig M, et al. Repair of long-bone fractures in cats and small dogs with the Unilock mandible locking plate system. Vet Comp Orthop Traumatol 2009;22:398–405.
7. Haaland PJ, Sjöström L, Devor M, et al. Appendicular fracture repair in dogs using the locking compression plate system: 47 cases. Vet Comp Orthop Traumatol 2009;22:309–315.
8. Nicetto T, Petazzoni M, Urizzi A, et al. Experiences using the Fixin locking plate system for the stabilization of appendicular fractures in dogs: a clinical and radiographic retrospective assessment. Vet Comp Orthop Traumatol 2013;26:61–68.
9. Barnhart MD, Rides CF, Kennedy SC, et al. Fracture repair using a polyaxial locking plate system (PAX). Vet Surg 2013;42:60–66.
10. Vallefuoco R, Le Pommellet H, Savin A, et al. Complications of appendicular fracture repair in cats and small dogs using locking compression plates. Vet Comp Orthop Traumatol 2016;29:46–52.
11. Johnson SJ, von Pfeil DJF, Déjardin LM, et al. Internal fracture fixation. In: Johnson SJ, Tobias KM, eds. Veterinary surgery small animal. 2nd ed. St Louis: Elsevier, 2018;654–690.
12. Boero Baroncelli A, Reif U, Bignardi C, et al. Effect of screw insertion torque on push-out and cantilever bending properties of five different angle-stable systems. Vet Surg 2013;42:308–315.
13. Hurt RJ, Syrcle JA, Elder S, et al. A biomechanical comparison of unilateral and bilateral String-of-Pearls™ locking plates in a canine distal humeral metaphyseal gap model. Vet Comp Orthop Traumatol 2014;27:186–191.
14. Kim SE, Lewis DD. Corrective osteotomy for procurvatum deformity caused by distal femoral physeal fracture malunion stabilised with String-of-Pearls locking plates: results in two dogs and a review of the literature. Aust Vet J 2014;92:75–80.
15. Ness MG. Repair of Y-T humeral fractures in the dog using paired ‘String of Pearls’ locking plates. Vet Comp Orthop Traumatol 2009;22:492–497.
16. Early P, Mente P, Dillard S, et al. In vitro biomechanical evaluation of internal fixation techniques on the canine lumbosacral junction. PeerJ 2015;3:e1094.
17. Demianiuk RM, Benamou J, Rutherford S, et al. Effect of screw type and distribution on the torsional stability of 3.5 mm string of pearls locking plate constructs. Vet Surg 2015;44:119–125.
18. Goh CSS, Santoni BG, Puttlitz CM, et al. Comparison of the mechanical behaviors of semicontoured, locking plate-rod fixation and anatomically contoured, conventional plate-rod fixation applied to experimentally induced gap fractures in canine femora. Am J Vet Res 2009;70:23–29.
19. Boulton CL, Kim H, Shah SB, et al. Do locking screws work in plates bent at holes? J Orthop Trauma 2014;28:189–194.
20. Ioannou C, Knight M, Daniele L, et al. Effectiveness of the surgical torque limiter: a model comparing drill- and hand-based screw insertion into locking plates. J Orthop Surg Res 2016;11:118.
21. Tremolada G, Lewis DD, Paragnani KL, et al. Biomechanical comparison of a 3.5-mm conical coupling plating system and a 3.5-mm locking compression plate applied as plate-rod constructs to an experimentally created fracture gap in femurs of canine cadavers. Am J Vet Res 2017;78:712–717.
22. Chao P, Conrad BP, Lewis DD, et al. Effect of plate working length on plate stiffness and cyclic fatigue life in a cadaveric femoral fracture gap model stabilized with a 12-hole 2.4 mm locking compression plate. BMC Vet Res 2013;9:125.
23. Field EJ, Parsons K, Etches JA, et al. Effect of monocortical and bicortical screw numbers on the properties of a locking plate-intramedullary rod configuration. An in vitro study on a canine femoral fracture gap model. Vet Comp Orthop Traumatol 2016;29:459–465.
