Abstract
Purpose
To compare the biomechanical properties of three plate stabilization techniques for midshaft clavicle fractures: anatomical bicortical locking construct, anatomical unicortical locking construct, and reconstruction bicortical locking construct.
Methods
We analyzed superior plating of the clavicle using an anatomical clavicle plate (Acumed) with both bicortical or unicortical screw fixation and a locking reconstruction plate (DePuy-Synthes). Twenty-one fourth generation composite clavicles were used for non-destructive stiffness testing in axial loading, bending, and torsion. Fifteen composite clavicles and 17 foam clavicles were used for cyclic failure testing using a combined loading method that included all three loading modes.
Results
No significant differences were found between the three constructs in torsional stiffness nor in vertical bending loading. In axial loading, the anatomical bicortical locking construct was significantly stiffer than either anatomical unicortical locking construct or the reconstruction bicortical locking construct. The unicortical fixation was also significantly stiffer than the reconstruction bicortical locking construct. Regarding failure testing, there was not a significant difference between the bicortical and unicortical anatomical locking constructs; however, both were significantly stronger than bicortical screw fixation in the reconstruction plate.
Conclusion
Based on the biomechanical performance of these constructs, unicortical locked plate fixation may be a reasonable option in the treatment of displaced midshaft clavicle fracture fixation.
Keywords: clavicle, fracture, unicortical, biomechanical
Introduction
In recent years, there has been an exponential growth in the surgical treatment of midshaft clavicle fractures.1This is due to the recognition of fracture patterns associated with unsatisfactory results such as comminuted fractures, those with a displacement > 1.5 to 2 cm, and those with a third interposed vertical fragment.2,3In this subgroup of fractures, several clinical trials have shown that, compared to the non-surgical treatment, stabilization with plates and screws results in better functional results, a faster return to activities of daily living and sports, and a lower rate of major complications such as nonunion and malunion.4–7
It is also well known that the use of locking screws provides superior stabilization and healing potential compared to standard non-locking screws for plate osteosynthesis, particularly in poor quality bone when there is comminution preventing fracture site compression.8,9
However, surgical stabilization with plates and screws is not exempt from complications. The most frequent ones are those related to the implant, infection, and wound-related problems.10Yet, the most feared complications derived from the osteosynthesis using plates and screws are the neurovascular ones, including direct injury to subclavian vessels, deep vein thrombosis of the subclavian vein or of any of its branches, and direct compression of these structures caused by the screws.11–15
One of the strategies proposed to reduce the risk of damaging the neurovascular structures located below the clavicle consists of fixing the plate with unicortical locked screws. This technique has two main advantages. First, it would not be necessary to drill the inferior cortex of the clavicle, thus reducing the risk of injury of subclavian anatomic structures when drilling pilot holes for the bicortical screws. Second, by preventing the tip of the screw from piercing through the inferior cortex, the direct injury that the screw might produce by compression or tearing of the subclavian neurovascular structures is avoided.16–20Although osteosynthesis with plates and unicortical screws seems to be a promising technique, the literature remains unclear as to whether this fixation is stable enough to guarantee the necessary initial stability for bone healing.20
The aim of this study was to compare the biomechanical properties of three plate stabilization techniques for midshaft clavicle fractures: anatomical bicortical locking construct (ABLC), anatomical unicortical locking construct (AULC), and reconstruction bicortical locking construct (RBLC).
Methods
Clavicle models
Biomechanical testing was performed on both high and low bone quality surrogate models. These were represented by fourth generation high strength composite clavicles and low-density foam synthetic clavicle models (Sawbones Models 3408-1 and 1020, respectively, Pacific Research Laboratories, Vashon Island, WA). All specimens were of the same size (large) and were anatomically left clavicles. Every specimen was fixed at each end centered in a short (5 cm length by 5 cm diameter) section of PVC pipe using epoxy resin body filler (Bondo, 3M, St Paul, MN). Each plate was applied first at the superior aspect of the intact clavicle to ensure that the positioning was anatomic and consistent between specimens. The 3.5 locking reconstruction plates (DePuy-Synthes) were applied with three bicortical locking screws on each side of the fracture site (RBLC). The anatomical locking plates (Acumed TM, Hillsboro, OR) were applied with either three bicortical screws on each side of the fracture site (ABLC) and three unicortical screws on each side of the fracture site (AULC). In the anatomical plates, locking screws were placed in the first, third, fourth, and sixth holes and standard screws were placed in the second and fifth holes. Next, all of the clavicles were osteotomized between the third and fourth screws, creating an 8 mm gap representing an unstable comminuted midshaft fracture. Finally, because the synthetic bone models had no canal, we removed the plates and screws and the canal was drilled out in each direction from the fracture site (Figure 1). Then the plates and screws were replaced.
Figure 1.
Photograph of a fourth generation composite clavicle with drill out intramedularry canal showing the unicortical screw placement.
Stiffness testing
A total of 21 composite clavicles (seven per group: n = 7) were used for non-destructive stiffness testing in axial loading, bending, and torsion.
Axial loading: PVC caps were placed on the pipe sections at each end of the clavicle specimens and connected to the servohydraulic load frame (MTS model 858 Bionix, Eden Prairie, MN) using universal joints. A compressive displacement was applied at a rate 0.1 mm/s and the applied load and displacement were recorded. The slope of the steepest linear portion of the load–displacement curve was used to calculate the axial stiffness for each test (Figure 2).
Figure 2.
Photograph showing the biomechanical evaluation of the axial loading the clavicle.
Torsional loading: The same fixtures that were used in the axial tests were used to apply a rotational displacement while angle and torque were recorded. The slope of the steepest linear portion of the torque–angle curve was used to calculate the torsional stiffness for each test.
Posterior bending loading: Three-point (cantilever) bending was applied to each construct by mounting the universal joint from the sternal end of the mounted clavicle to the base of the load frame with a rigid support under the diaphysis between the second and third screws. A downward vertical displacement was applied to the acromial end of the clavicle at a distance of 9 cm from the fracture gap. The clavicle was oriented for this test as if the force on the acromial end was in a posterior direction. Because the plates were placed on the superior surface of the clavicles, this loading caused bending in the plane of the plate (Figure 3). The slope of the steepest linear portion of the load–displacement curve was used to calculate the bending stiffness for each test.
Figure 3.
Photograph showing the biomechanical evaluation of the posterior bending of the clavicle.
Vertical bending loading: Three-point (cantilever) bending was applied to each construct by mounting the universal joint from the sternal end of the mounted clavicle to the base of the load frame with a rigid support under the diaphysis between the second and third screws. A downward vertical displacement was applied to the acromial end of the clavicle at a distance of 9 cm from the fracture gap. The clavicle was oriented for this test as if the force on the acromial end was in the inferior or downward direction (Figure 4). Because the plates were placed on the superior surface of the clavicles, this loading caused bending in the plane perpendicular to the plane of the plate. The slope of the steepest linear portion of the load–displacement curve was used to calculate the bending stiffness for each test.
Figure 4.
Photograph showing the biomechanical evaluation of the vertical bending of the clavicle.
Failure testing
Cyclic failure: Cyclic failure testing was performed on both composite clavicle specimens and low-density foam clavicle specimens as combined loading that included axial, bending, and torsional loading applied in a step wise increasing fashion. The specimens were mounted in the load frame using a 34 mm offset plate before attaching the universal joints, such that an axial displacement would generate both axial loading and bending of the clavicle. A torsional load was simultaneously applied along with the axial displacement which resulted in the following loading protocols.
For the composite bones: Applied cyclic torque began at 0.6 N m and rose to a maximum of 6 N m in 100 cycle increments over 10 loading intervals. Simultaneously applied cyclic axial load started at 10 N and rose to 80 N in 100 cycle increments over the same 10 loading intervals. In this way, the applied bending started at 0.34 N m and rose to 2.7 N m in 100 cycle increments over the 10 loading intervals. The number of specimens per group for composite bone testing was RBLC: 4, ABLC: 6, AULC: 5.
For the foam bones: Applied cyclic torque began at 0.2 N m and rose to a maximum of 2.0 N m in 100 cycle increments over 10 loading intervals. Applied cyclic axial load started at 15 N and remained between 15 and 20 N throughout 10 loading intervals. Bending started at 0.5 N m and remained between 0.5 and 0.7 N m throughout 10 loading intervals. The number of specimens per group for foam bone testing was ABLC: 10 and AULC: 7. Cyclic failure tests were not done for the locking reconstruction group.
Cyclic failure testing of the composite bones was intended to test the strength of the plate and screw constructs because the bones were not expected to be the source of failure. Cyclic failure testing of the foam bones was intended to test the strength of the screw–bone interface, because in exceptionally poor quality bone as is found in severe osteoporosis, the screw–bone interface is the weakest part of the construct and the screws are susceptible to pullout failure. The sawbones foam bones thus may represent a worst case scenario for screw-plate constructs in poor quality bone.
All cyclic failure results are presented as the torsional value at failure. This is because the failure mode was always observed to be in torsional, whether it was twisting of the plate, a spiral bone fracture, or splitting and pullout of the screws.
Torsion failure:A simple torsion to failure test was performed on eight composite bone samples that were left over after stiffness testing. There were five ABLC and three AULC. They were set up in the same way as the torsional stiffness testing. A constantly increasing rotational displacement was applied and the torque was simultaneously recorded. The maximum torque achieved before failure was recorded.
Statistical analysis
When the testing results for three groups were compared, single-factor analysis of variance was used followed by post hoc t-tests to compare between individual groups. In the cases when only two groups were tested, comparisons were made by Student’s t-test. Statistical significance was set at p < 0.05 in all cases.
Results
Stiffness testing
Axial loading:All three of the plated clavicle constructs were quite stiff in axial loading. The ABLC (205 ± 52.9 N/mm) was significantly stiffer than either the AULC (145 ± 34.1 N/mm) or the RBLC (90 ± 6.0 N/mm). The unicortical locking construct was also significantly stiffer than the recon locking construct (Figure 5).
Figure 5.
Axial loading stiffness.
Torsional loading:In pure torsional stiffness testing, there was little difference between ABLC (0.43 ±0.03 N m/deg), AULC (0.41 ± 0.03 N m/deg), or the RBLC (0.37 ± 0.07 N m/deg) (Figure 6).
Figure 6.
Torsional loading stiffness.
Posterior bending loading:Both the ABLC (22.6 ± 4.7 N/mm) and the AULC (22.1 ± 3.1 N/mm) were significantly stiffer than the RBLC (17.3 ±3.2 N/mm). The posterior direction was also significantly stiffer than the vertical direction because of the orientation of the plates (p < 0.001) (Figure 7).
Figure 7.
Posterior bending loading.
Vertical bending loading:In the vertical direction, the ABLC (9.1 ± 0.98 N/mm), the AULC (8.8 ± 1.33 N/mm), and the RBC (7.7 ± 0.66 N/mm) were not significantly different from each other (p > 0.05) (Figure 8).
Figure 8.
Vertical bending loading.
Failure testing
Cyclic failure (composite):Cyclic testing of the composite clavicle constructs showed very little difference between the ABLC (6.0 ± 0.0 N m) and unicortical locking construct (5.7 ± 0.7 N m) but both were statistically significantly stronger than the recon locking construct (3.3 ± 0.3 N m) (Figure 9).
Figure 9.
Cyclic failure (composite).
Cyclic failure (foam):The ABLC (1.5 ± 0.2 N m) group survived significantly higher torsional load cycles compared to the AULC (1.0 ± 0.3 N m) and the mode of failure was different. In the case of the ABLC, the foam bone fractured through a screw hole, whereas in the AULC, the screws began to pull out of the foam after the torque reached 1.0 N m. The RBLC was not tested in the foam bones because this test was specifically intended to determine any differences in screw (number of cortices) choice in poor quality bone.
Torsion failure:The torque to failure strength of ABLC (11.4 ± 3.1 N m) was significantly stronger than the AULC (7.4 ± 0.1 N m). The failure of the unicortical screw constructs appeared to occur as a result of greater stress at the screw–bone interface. However, the strength of AULC (7.4 N m) is far greater than any loads that would be expected in vivo on the clavicle. From the cyclic data, we can estimate the torsional strength of the recon locking construct to be approximately 3.3 N m.
Discussion
The main findings of this study were that fractures of the clavicle fixed with ABLC were significantly more stable fixation with anatomical unicortical or reconstruction locking constructs regarding axial loading stiffness, vertical bending stiffness, and torsional strength to failure. However, there were not significant differences between the three fixation methods regarding torsional loading stiffness nor posterior bending stiffness.
The biomechanical properties between bicortical and unicortical plate fixation of clavicular fractures have previously been compared by other authors with conflicting results.16–19Hamman et al.20evaluated the strength of unicortical locked plating versus traditional bicortical non-locking fixation methods. In their study, the authors randomized 10 matched pairs of fresh, frozen cadaver clavicle fractures to receive fixation using precontoured plates with either standard bicortical non-locking screws or unicortical locking screws. The authors found that there was no significant difference between the constructs in terms of axial stiffness or load to failure. However, rotational stiffness varied significantly (locking 1.70 ± 0.91 N m/mm, non-locking 2.49 ± 0.78 N m/mm, p = 0.049) with bicortical non-locking constructs exhibiting higher torque values. The authors concluded that unicortical locking screw fixation provided a suitable biomechanical environment for bone healing. Despite this, an important limitation of this study was that bicortical locking screws were not tested. It would have been useful to have a comparative group of clavicles fixed with bicortical locking screws since this could have yielded different results. In a similar study, Renfree et al.19also compared clavicle fixation with precontoured clavicle plates with unicortical locking and bicortical non-locking screws but added a third group stabilized with intramedullary Rockwood clavicle pins. Similarly, to Hamman et al., the authors found that both locking and non-locking constructs provided similar rigid fixation for clavicular fractures. However, compared to plate fixation, intramedullary pin fixation appeared to be inadequate where rotational stiffness is required.
Brayman et al.16compared the biomechanical properties of unicortical and bicortical locking screws fixation in precontoured versus manually contoured locking clavicle plates in 48 sawbone composite human clavicle specimens. The authors reported no significant differences between unicortical and bicortical fixation in failure load, cantilever bending, and cross body stiffness. However, bicortical fixation was significantly stiffer than unicortical fixation in torsion when the same type of plate was compared. The authors concluded that unicortical locked plate fixation may be a reasonable option in the treatment of displaced midshaft clavicle fracture fixation.
Overturf et al.21proposed an alternative to avoid penetrating the inferior cortex of the clavicle which they called unicortical far-cortex abutting screw. This screw contacts the distal cortex without penetrating, effectively improving positional stability compared with traditional unicortical fixation by increasing the length of screw in contact with bone. Croley et al.17in a biomechanical cadaveric study assessed the biomechanical properties of bicortical locking screws, unicortical locking screws, and unicortical far-cortex-abutting locking screw fixation in a cadaver model of comminuted midshaft clavicle fractures stabilized with a locking plate placed on the superior surface of the clavicle. Nine pairs of adult fresh-frozen cadaver clavicles were allocated into the three groups. Both the bicortical and the unicortical far-cortex-abutting constructs were significantly stiffer than the unicortical construct in cyclical torsion and torsional failure stiffness. There was no difference between bicortical and unicortical far-cortex-abutting for torsional failure stiffness. The authors concluded that unicortical far-cortex-abutting locking screw fixation provides comparable mechanical properties under axial and torsional loads to bicortical fixation, without penetrating the far cortex. Nonetheless, although this technique obviates far cortex penetration, and thereby protects nearby anatomical structures, it is technically very complex to drill the inferior cortex only partially so that the screw is locked in it but does not pierce through it. In addition, injury to neighboring anatomic structures may also occur when drilling the inferior cortex and not only as a result of the screws sticking out of it.
An important finding of our study was that we found no significant differences between plate fixation using bicortical or unicortical locking screws regarding torsional loading stiffness, and posterior nor vertical bending stiffness. It is important to note that the initial movements performed by the patient during rehabilitation, such as forward flexion and internal and external rotation, mainly involve torsional and bending forces.22Therefore, it is very likely that both unicortical and bicortical fixations will be sufficient to achieve the necessary stability for bone healing of clavicle fractures. Although we found a significantly stronger axial loading stiffness with bicortical locking screws, the axial load in the immediate postoperative stage is mild and it is not very likely to constitute a relevant clinical advantage.
Finally, another interesting finding of this study was that the ABLC (1.5 ± 0.2 N m) group survived significantly higher torsional load cycles compared to the unicortical locking construct (1.0 ± 0.3 N m) in the foam bone constructs. This could be relevant in the clinical setting due to the fact that foam bone represents very poor quality osteoporotic clavicles. Moreover, the failure modes were different between both groups. The clavicles fractured through a screw tract in the bicortical group, and mostly the screws pulled out of the bone in the unicortical group. These results suggest that in older patients, plate fixation with bicortical locking screws may be the best option.
Our study has several limitations. It should be noted that, as in every biomechanical study, this work only evaluates what happens at time zero after stabilization. However, the postoperative period of surgical stabilization of clavicle fractures includes one month of immobilization with a sling. During this period, bone callus formation progresses and, therefore, as the bone healing process progresses, part of the load is absorbed by the bone and thus the necessary stiffness required from the implant is lower. Another important limitation is that up till now it remains unclear which degree of construct stiffness or strength is most optimal to stimulate adequate bone healing or which loading thresholds fixation constructs have to withstand in vivo. Therefore, we do not know if the statistically significant differences reported in our study are really clinically relevant. Future studies comparing the plate fixation of displaced midshaft clavicular fractures with bicortical and unicortical screws are necessary to confirm whether the advantages of unicortical screws are reproduced in the clinical setting.
Conclusion
Unicortical locked plate fixation may be a reasonable option in the treatment of displaced midshaft clavicle fracture fixation to avoid complications associated with inferior hardware penetration following clavicle fracture fixation based on the biomechanical performance of these constructs.
Acknowledgement
The authors acknowledge the support of the University of Louisville, Kentucky (USA).
Footnotes
Declaration of Conflicting Interests: The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Zaidenberg, Voor, and Rossi have no relevant financial relationships to disclose. Pereira is a paid consultant for Acumed TM, Hillsboro, OR.
Ethical Approval and Informed Consent: Not applicable.
Funding: The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors have received a grant from Acumed to complete the study.
Guarantor: EEZ.
ORCID iDs
Ezequiel E Zaidenberg https://orcid.org/0000-0002-1535-0586
Luciano A Rossi https://orcid.org/0000-0002-1397-2402
References
- 1.Wiesel B, Nagda S, Mehta S, et al. Management of midshaft clavicle fractures in adults. J Am Acad Orthop Surg 2018; 26: 468–476. [DOI] [PubMed] [Google Scholar]
- 2.Vautrin M, Kaminski G, Barimani B, et al. Does candidate for plate fixation selection improve the functional outcome after midshaft clavicle fracture? A systematic review of 1348 patients. Shoulder Elbow 2019; 11: 9–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Woltz S, Stegeman SA, Krijnen P, et al. Plate fixation versus nonoperative treatment for displaced midshaft clavicular fractures: a multicenter randomized controlled trial. J Bone Joint Surg Am 2017; 99: 106–112. [DOI] [PubMed] [Google Scholar]
- 4.Woltz S, Krijnen P, Schipper IB. Plate fixation versus nonoperative treatment for displaced midshaft clavicular fractures: a meta-analysis of randomized controlled trials. J Bone Joint Surg Am 2017; 99: 1051–1057. [DOI] [PubMed] [Google Scholar]
- 5.Qvist AH, Vaesel MT, Jensen CM, et al. Plate fixation compared with nonoperative treatment of displaced midshaft clavicular fractures: a randomized clinical trial. Bone Joint J 2018; 100-B: 1385–1391. [DOI] [PubMed] [Google Scholar]
- 6.Bhardwaj A, Sharma G, Patil A, et al. Comparison of plate osteosynthesis versus non-operative management for mid-shaft clavicle fractures: a prospective study. Injury 2018; 49: 1104–1107. [DOI] [PubMed] [Google Scholar]
- 7.Xu J, Xu L, Xu W, et al. Operative versus nonoperative treatment in the management of midshaft clavicular fractures: a meta-analysis of randomized controlled trials. J Shoulder Elbow Surg 2014; 23: 173–181. [DOI] [PubMed] [Google Scholar]
- 8.DeKeyser GJ, Kellan PJ, Haller JM. Locked plating and advanced augmentation techniques in osteoporotic fractures. Orthop Clin North Am 2019; 50: 159–169. [DOI] [PubMed] [Google Scholar]
- 9.Hak DJ, Banegas R, Ipaktchi K, et al. Evolution of plate design and material composition. Injury 2018; 49: S8–S11. [DOI] [PubMed] [Google Scholar]
- 10.Navarro RA, Gelber JD, Harrast JJ, et al. Frequency and complications after operative fixation of clavicular fractures. J Shoulder Elbow Surg 2016; 25: 125–129. [DOI] [PubMed] [Google Scholar]
- 11.Singh A, Schultzel M, Fleming JF, et al. Complications after surgical treatment of distal clavicle fractures. Orthop Traumatol Surg Res 2019; 105: 853–859. [DOI] [PubMed] [Google Scholar]
- 12.Adla DN, Ali A, Shahane SA. Upper-extremity deep-vein thrombosis following a clavicular fracture. Eur J Orthop Surg Traumatol 2004; 14: 177–179. [DOI] [PubMed] [Google Scholar]
- 13.Lim EV, Day LJ. Subclavian vein thrombosis following fracture of the clavicle: a case report. Orthopedics 2017; 10: 349–351. [DOI] [PubMed] [Google Scholar]
- 14.Rossi LA, Piuzzi NS, Bongiovanni SL, et al. Extrinsic subclavian vein compression after osteosynthesis of a midshaft clavicular fractures in an athlete. Case Rep Orthop 2015; 2015: 981293–981293. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lewis SD, Chew FS. Clavicle fixation screw impingement causing subclavian artery pseudoaneurysm. Radiol Case Rep 2019; 14: 1148–1150. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Brayman JT, Taylor ML, Baldini T, et al. Unicortical versus bicortical locked plate fixation in midshaft clavicle fractures. Orthopedics 2015; 38: 411–416. [DOI] [PubMed] [Google Scholar]
- 17.Croley JS, Morris RP, Amin A, et al. Biomechanical comparison of bicortical, unicortical, and unicortical far-cortex-abutting screw fixations in plated comminuted midshaft clavicle fractures. J Hand Surg Am 2016; 41: 703–711. [DOI] [PubMed] [Google Scholar]
- 18.Hulsmans MH, van Heijl M, Houwert RM, et al. Surgical fixation of midshaft clavicle fractures: a systematic review of biomechanical studies. Injury 2018; 49: 753–765. [DOI] [PubMed] [Google Scholar]
- 19.Renfree T, Conrad B, Wright T. Biomechanical comparison of contemporary clavicle fixation devices. J Hand Surg Am 2010; 35: 639–644. [DOI] [PubMed] [Google Scholar]
- 20.Hamman D, Lindsey D, Dragoo J. Biomechanical analysis of bicortical versus unicortical locked plating of mid-clavicular fractures. Arch Orthop Trauma Surg 2011; 131: 773–778. [DOI] [PubMed] [Google Scholar]
- 21.Overturf SJ, Morris RP, Gugala Z, et al. Biomechanical comparison of bicortical locking versus unicortical far-cortex abutting locking screw-plate fixation for comminuted radial shaft fractures. J Hand Surg Am 2014; 39: 1907–1913. [DOI] [PubMed] [Google Scholar]
- 22.Iannolo M, Werner FW, Sutton LG, et al. Forces across the middle of the intact clavicle during shoulder motion. J Shoulder Elbow Surg 2010; 19: 1013–1017. [DOI] [PubMed] [Google Scholar]