Clinical Biomechanics 30 (2015) 405–410
Contents lists available at ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
Angular stable lateral plating is a valid alternative to conventional plate fixation in the proximal phalanx. A biomechanical study R. Shanmugam a,b,1, M. Ernst a,⁎,1, K. Stoffel c,d, M.F. Fischer a, D. Wahl a, R.G. Richards a, B. Gueorguiev a a
AO Research Institute Davos, Davos, Switzerland Orthopaedic Department, University of Malaya, Kuala Lumpur, Malaysia Cantonal Hospital Baselland, Liestal, Switzerland d University of Basel, Basel, Switzerland b c
a r t i c l e
i n f o
Article history: Received 8 September 2014 Accepted 18 March 2015 Keywords: Proximal phalanx Lateral plating Fracture fixation
a b s t r a c t Background: Dorsal plating is commonly used in proximal phalanx fractures but it bears the risk of interfering with the extensor apparatus. In this study, dorsal and lateral plating fixation methods are compared to assess biomechanical differences using conventional 1.5 mm non-locking plates and novel 1.3 mm lateral locking plates. Methods: Twenty-four fresh frozen human cadaveric proximal phalanges were equally divided into four groups. An osteotomy was set at the proximal metaphyseal–diaphyseal junction and fixed with either dorsal (group A) or lateral (group B) plating using a 1.5 mm non-locking plate, or lateral plating with a novel 1.3 mm locking plate with bicortical (group C) or unicortical (group D) screws. The specimens were loaded in axial, dorsovolar and mediolateral direction to assess fixation stiffness followed by a cyclic destructive test in dorsovolar loading direction. Findings: Axial stiffness was highest in group D (mean 321.02, SEM 21.47 N/mm) with a significant difference between groups D and B (P = 0.033). Locking plates (groups C and D) were stiffer than non-locking plates under mediolateral loading (P = 0.007), no significant differences were noted under dorsovolar loading. Furthermore, no significant differences were observed under cyclic loading to failure between any of the study groups. Interpretation: No considerable biomechanical advantage of using a conventional 1.5 mm dorsal non-locking plate was identified over the novel 1.3 mm lateral locking plate in the treatment of proximal phalanx fractures. Since the novel low-profile plate is less disruptive to the extensor mechanism, it should be considered as a valid alternative. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Phalangeal fractures are among the commonest hand fractures constituting about 10% of all hand injuries (Kamath et al., 2011; Ouellette et al., 2004). Although a lot of phalangeal fractures can be managed non-operatively, operative treatment may be indicated in the presence of shortening, excessive deformity, involvement of multiple rays, open fractures or failure to achieve acceptable reduction (Adams et al., 2013; Wong et al., 2008; Kamath et al., 2011; Nunley and Kloen, 1991; Ouellette et al., 2004). Following hand fractures, early physiotherapy is important to regain a good functional outcome (Hardy, 2004; Lu et al., 1996; Nunley and Kloen, 1991; Page and Stern, 1998). This requires a method of fixation rigid enough to allow early commencement of physiotherapy without causing any loss of reduction or excessive movement
⁎ Corresponding author at: AO Research Institute Davos, Clavadelerstrasse 8, 7270 Davos Platz, Switzerland. E-mail address: [email protected] (M. Ernst). 1 These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.clinbiomech.2015.03.019 0268-0033/© 2015 Elsevier Ltd. All rights reserved.
at the fracture site that may be related to pain and/or non-unions (Freeland et al., 2003). Plating can hold an accurate fracture reduction rigidly and is suitable for fixation of phalangeal fractures as it would allow for early physiotherapy. There are two well described techniques for phalanx plating: the dorsal and the lateral technique (Berman et al., 1999; Hattori et al., 2007; Haughton et al., 2012; Ruchelsman et al., 2010). In dorsal plating, the incision is made over the dorsal aspect of the finger splitting the extensor tendon to access the bone. The plate is then placed on the dorsal aspect of the bone just underneath the extensor tendon. This often causes irritation during movement and might even lead to disruption of the tendon (Pun et al., 1991; Stern et al., 1987). The most frequent complication observed after open reduction and plate fixation of the phalanx is reduced mobility of the extensor tendons caused by scar tissue formation and resulting in a stiff joint (Kurzen et al., 2006; Page and Stern, 1998). In addition, in case of too long screws, their tips on the volar aspect might irritate the flexor tendon or even cause its rupture (Fambrough and Green, 1979; Kurzen et al., 2006). On the other hand, lateral plating carries a lower risk of irritation to the extensor mechanism as the respective tendon does not need to be split so that
406
R. Shanmugam et al. / Clinical Biomechanics 30 (2015) 405–410
scarring and adhesion are potentially avoided (Dabezies and Schutte, 1986; Kozin et al., 2000). Nevertheless, since the finger is a very compact structure, there is still some risk for the implant to be prominent or impinge on the neurovascular structures in close proximity or at the far cortex in case of bicortical screw fixation. Therefore, it is important to minimize surgical trauma during plating for good clinical outcomes applying a less invasive technique with low-profile implants in order to reduce possible irritation to surrounding tissues. Rigidity of various fixation methods was investigated in previous studies (Afshar et al., 2012; Black et al., 1986; Lu et al., 1996; Nunley and Kloen, 1991; Ouellette et al., 2004; Prevel et al., 1995). Most of them were conducted applying quasistatic loading, which is less representative for the repetitive movements encountered during rehabilitation, or three-point bending of the fingers with limited physiological relevance. In addition, there is no consensus regarding the predominant loading type of the fingers during rehabilitation following a phalanx fracture (An et al., 1985; Berme et al., 1977; Wu et al., 2008). In spite of this and the paucity of supporting data, dorsal plating is widely used for phalangeal fracture fixation due to its straightforward approach and the assumption that thereby a stabilizing tension band effect is achieved (Black et al., 1986; Prevel et al., 1995). One of the most commonly used implants for phalanx fractures is the T-plate (DePuy Synthes, Solothurn, Switzerland), available as 1.5 mm non-locking and locking plate, that can be used for both dorsal and lateral application by trimming and shaping the plate accordingly. Recently, a novel, anatomically shaped low-profile locking plate (DePuy Synthes, Solothurn, Switerzland, not yet available) has been designed especially for application on the lateral aspect of the proximal phalanx. The aim of this study was to compare the biomechanical performance of the novel low-profile lateral locking plate for proximal phalanx fracture fixation versus an established non-locking plate routinely used in clinics. We hypothesized that with the use of the locking screw technology a low-profile implant can reach comparable fixation rigidity as a bulkier conventional non-locking plate. Furthermore, it is expected that lateral plating is not biomechanically inferior to dorsal plating under consideration of various loading conditions. 2. Methods Twenty-four fresh frozen (−20 °C) human proximal phalanx cadaveric specimens from the 2nd to 4th fingers were used in this study. The specimens were thawed in a fridge (4 °C) for 24 h prior to preparation and biomechanical testing. Testing was performed at room temperature with phalanges kept moist. All specimens were subjected to bone mineral density (BMD) measurement using Xtreme CT (Scanco Medical, Brüttisellen, Switzerland) at a resolution of 82 μm. The region of interest was selected in the whole proximal phalanx from the proximal to the distal condyles encompassing the cortical and cancellous components. The mean computed value was considered as BMD of the particular phalanx. The 24 phalanges were equally divided into four study groups (Table 1) with no significant difference in the BMD distribution between the groups. Length of all specimens was set to 48 mm by cutting a part of the distal condyle in order to standardize the moment arm during the following cantilever bending tests. A transverse osteotomy was cut freehand 11 mm distally to the proximal joint line using an oscillating saw (Colibri system, DePuy Synthes, Solothurn, Switzerland) to simulate a
transverse proximal fracture at the metaphyseal–diaphyseal junction. The specimens were then subjected to fixation using two types of plates with different fixation methods as described in Table 1. All plates were trimmed to equalize the length and contoured to the bone as needed. In all study groups the proximal and the distal fragments were fixed with two screws each (Fig. 1). The fracture reduction was performed under slight compression and the fragments were fixed according to the manufacturer's guidelines (DePuy Synthes, Solothurn, Switzerland). Reproducibility of both fracture pattern and fixation technique with correct and consistent alignment and implants placement were ensured. K-wires were passed through both proximal and distal ends of each bone and incorporated into Polymethylmethacrylate (PMMA) (Troller AG, Fulenbach, Switzerland) to embed the specimens properly. Proximal and distal embedding was performed using custom made molds allowing the embedded specimens to fit into a jig during biomechanical testing (Fig. 2). The mold for embedding the distal end of the phalanx comprised a central opening at the bottom through which a metal pin was inserted before embedding. The specimen was then positioned with its distal end within the mold and the long bone axis aligned along the pin axis. The pin was incorporated in the PMMA block after embedding. All samples were prepared by a single surgeon (R.S.). Biomechanical testing was performed using a Bose 3220 material testing machine (BOSE, Eden Prairie, MN, USA) with a 225 N load cell coupled to a vertical slide with negligible friction. The specimens were mounted on a custom made jig allowing positioning in multiple directions according to the type of required loading (Fig. 3). For specimen's axial compression, the load was introduced via a beam connected to the vertical slide. For dorsovolar (DV) and mediolateral (ML) cantilever bending, the pin was inserted for loading into a hole in the vertical slide, keeping a 45-mm lever arm for each specimen. The load cell was coupled to the vertical slide. All specimens were firstly subjected to non-destructive axial compressive loading, followed by DV and ML cantilever bending to mimic early rehabilitation range-of-motion (ROM) exercises (Hardy, 2004; Nunley and Kloen, 1991). Finally, a destructive cyclic test in DV cantilever bending was run. Each of the three nondestructive tests comprised three successive quasistatic ramps at a rate of 0.01 mm/s. In case of axial compressive loading, each ramp ranged from 1 N (valley) to 5 N (peak) (Fig. 3A). During the nondestructive cantilever bending tests (Fig. 3B and C) the load was applied symmetrically at 0 N mean level and 3.0 N amplitude, except for the test phase when the specimen was loaded with the implant on the compression side. In this case, the force was limited to 1.5 N to avoid excessive opening of the osteotomy at the far cortex. The destructive sinusoidal cyclic test was performed at 0.5 Hz in DV loading direction until bone-implant construct failure. The valley load was kept at a constant level of 1 N throughout the whole test while the peak load, starting at 5 N, was progressively increased by 0.007 N every cycle. Based on the machine data acquired from the load-cell and system's transducer at a rate of 50 Hz, the following parameters of interest were evaluated. Stiffness of the bone-plate construct in all loading directions was calculated from the linear range of the respective loaddisplacement curve of the last of the three quasi-static nondestructive ramped cycles, the first two ramped cycles served as preconditioning of the specimen. Moreover, plastic deformation after 200 and 400 sinusoidal cycles was calculated as the relative axial
Table 1 Study groups. Details of the four study groups indicating the type of implant and fixation method applied.
Implant Plate position Screws
Group A (n = 6)
Group B (n = 6)
Group C (n = 6)
Group D (n = 6)
T-adaptation plate Dorsal 1.5 mm non-locking Bicortical
T-adaptation plate Lateral 1.5 mm non-locking Bicortical
Phalangeal base plate Lateral 1.3 mm locking Bicortical
Phalangeal base plate Lateral 1.3 mm locking Unicortical
R. Shanmugam et al. / Clinical Biomechanics 30 (2015) 405–410
Fig. 1. Instrumentation details. The length of the two implants used in this study was standardized by cutting off the last 5 (T-adaptation plate) or 2 holes (Phalangeal base plate), respectively. Additionally, the 1st of the three perpendicular holes in the lateral nonlocking group was cut for lateral plating (B). The two groups instrumented with the novel lateral locking plate (C and D) differ only in the use of bicortical (C) or unicortical (D) screws. Solid line indicates the section where the plate was cut; dotted line indicates the location of the osteotomy.
displacement at valley load of 1 N with respect to its value at valley load of the fifth cycle, taken as reference. Finally, cycles to failure were defined as the number of cycles until reaching a plastic deformation of 3 mm, considered as an arbitrary criterion for clinical failure. Statistical analysis was performed using SPSS (Version 19, IBM SPSS, Chicago, IL, USA). Normal distribution and homogeneity of variances of all parameters of interest in the study groups were tested with Shapiro– Wilk and Levene tests, respectively. Univariate Analysis of Variance (ANOVA) with Bonferroni Post Hoc tests, Paired-Samples t-test, as well as General Linear Model (GLM) Repeated Measures with a Greenhouse–Geisser epsilon adjustment were applied to detect significant differences, setting the level of significance at P = 0.05 for all statistical tests. 3. Results BMD was not significantly different between the study groups (A: mean 342.27, SEM (standard error of mean) 50.85 mgHA/ccm, B: mean 329.91, SEM 42.99 mgHA/ccm, C: mean 363.13, SEM 20.92 mgHA/ccm, D: mean 333.27, SEM 45.86 mgHA/ccm). Stiffness values, calculated for the various loading conditions in terms of mean and SEM, are shown in Fig. 4. Under axial loading, stiffness values were generally higher in the locking plate constructs in comparison to the non-locking ones. Statistical significant difference under axial loading was detected between group B (non-locked lateral plating with bicortical screws, mean 256.09, SEM 10.00 N/mm) and D (locked lateral plating with unicortical screws, mean 321.02, SEM 21.47 N/mm), P = 0.033. No significant differences between the groups were observed for DV stiffness (A: mean 8.22, SEM 2.14 N/mm, B: mean 6.36, SEM
407
1.43 N/mm, C: mean 9.22, SEM 2.77 N/mm, D: mean 5.85, SEM 1.83 N/mm), in group A evaluated during the phase of the DV cantilever bending test with the plate on tension side. However, under ML loading, the constructs with locking plates (C: mean 18.32, SEM 4.73 N/mm, D: mean 16.57, SEM 4.97 N/mm) were significantly stiffer compared to the non-locked constructs (A: mean 6.86, SEM 1.42 N/mm, with plate on tension side, B: mean 6.44, SEM 0.93 N/mm), P = 0.007. There were no significant differences in stiffness noted between the dorsal and lateral non-locking plate groups in all modes of loading as well as between unicortical and bicortical locked plate constructs. Stiffness evaluation during the compression and extension phase of each DV and ML cantilever bending test in each group separately revealed significantly higher values when the plate was on the tension side (compression phase) in comparison to its position on the compression specimen side (extension phase), P = 0.004. There was a difference between ML and DV stiffness in the locking plate groups (average difference 9.1 N/mm, with P = 0.021 and 10.7 N/mm, with P N 0.05 for groups C and D, respectively) with the specimens under ML loading being stiffer. In contrast, the bending stiffness did not significantly differ between the loading directions in the non-locked groups. In the dorsal non-locked group, bending stiffness was on average 1.36 N/mm higher under DV loading, and in the lateral non-locked group 0.08 N/mm higher under ML loading. Plastic deformation progressively increased throughout the cyclic destructive test in all groups. After 200 cycles, the highest plastic deformation was observed in the dorsal non-locking group with 0.71 mm on average (SEM 0.23 mm), followed by locked lateral plating with unicortical screws (mean 0.58 mm, SEM 0.16 mm), locked lateral plating with bicortical screws (mean 0.54 mm, SEM 0.16 mm) and the lowest plastic deformation in the lateral non-locking group with 0.43 mm on average (SEM 0.18 mm). Similarly, the values for plastic deformation after 400 cycles were highest in the dorsal non-locking group (mean 0.99 mm, SEM 0.39 mm), lowest in the non-locking lateral plating group (mean 0.55 mm, SEM 0.22 mm) and the two lateral locking plate groups in between (mean 0.82, SEM 0.27 mm with bicortical screws; mean 0.76 mm, SEM 0.27 mm with monocortical screws). The plastic deformation increased significantly in all groups between 200 and 400 cycles (P = 0.022), differences between the groups were not statistically significant. Specimens of the dorsal non-locked group withstood the highest number of cycles to failure (mean 2350, SEM 688 cycles), followed by the locked lateral group with bicortical screws (mean 1983, SEM 433), locked lateral plating with monocortical screws (mean 1618, SEM 352) and non-locked lateral plating (mean 1493, SEM 386). However, no statistical significance was screened with regard to this parameter of interest. Typical catastrophic failure observed in all specimens was breakage of the bone bridge from the screw hole of the closest screw on the proximal fragment towards the osteotomy (Fig. 5). This resulted in loss of reduction due to the screw displacing along the fracture, thereby often pushing a small bone fragment into the osteotomy. 4. Discussion
Fig. 2. Exemplified specimen after embedding in PMMA, ready to be mounted for testing.
Early rehabilitation of operatively treated phalanx fractures typically comprises ROM exercises as well as tendon gliding exercises (Hardy, 2004). These are generally repetitive in nature and are both active and passive but under low load. Hence, a cyclic loading regime would be ideal for testing. At the same time very little data is available on the actual loads on the phalanx during such exercises. Thus, a test protocol which gradually increases the loading with each cycle was chosen. This is in contrast with other studies that use quasistatic or fatigue type cyclic loading (Black et al., 1986; Firoozbakhsh et al., 1993; Prevel et al., 1995). When applying plating as a method of fracture fixation, it is essential to place the plate on the tension side of the fractured bone (Egol et al., 2004; Ouellette et al., 2004). This enhances stability of fixation through
408
R. Shanmugam et al. / Clinical Biomechanics 30 (2015) 405–410
Fig. 3. Illustration of the test setup and loading modes with given force ranges during biomechanical testing. A shows a specimen mounted for axial loading, B and C show a specimen prepared for ML and DV cantilever bending, respectively. The specimen is rotated along the long bone axis to change between ML and DV loading, while the test setup remains the same. Arrows indicate the direction of loading.
the far cortical support under compression, taking up a substantial amount of load rather than depending on the plate entirely. This is evidenced in the current study by the significantly higher bending stiffness achieved during the compression phase of loading (plate on the tension side) compared to its extension phase (plate on compression side). In addition, this also implies the presence of good fracture reduction in the current study resulting in load sharing between bone and plate on tension side for all specimens. Dorsal plating with a non-locking T-plate was tested as a control group in our study because this implant is established for surgical fixation of proximal phalanx fractures and to the authors knowledge provides sufficient construct stability for functional aftercare. None of the other three test groups performed significantly worse than the control group, therefore, it is concluded that all four tested constructs provide the required stability during the rehabilitation period. This statement is supported by the fact that early rehabilitation primarily concentrates
on ROM exercises to prevent joint stiffness, thereby avoiding excessive loading that could lead to failure of the bone-implant construct. In our study the locking implants revealed higher stiffness compared to the non-locking ones in axial and mediolateral loading directions. This difference was highly significant in mediolateral cantilever bending but less obvious under axial compression, where the good fracture reduction with optimal cortical support probably prevented larger differences. This finding is consistent with similar outcomes of previous studies (Egol et al., 2004; Miller and Goswami, 2007) and may be explained with the angular stability of the locked screw-plate connection which does not allow for any slipping or change in the angulation so that screw loosening and pullout can be prevented. Consequently, it ultimately leads to a stiffer construct and higher pullout resistance. For that reason the novel low-profile plates tested in the current study are stiffer than the larger 1.5 mm non-locking plates under most of the loading conditions.
Fig. 4. Stiffness results. Comparison of mean values for the stiffness evaluated under DV and ML bending (left) and under axial compression (right). Error bars indicate standard error of mean, * indicate significant differences (*P b 0.05, **P b 0.01).
R. Shanmugam et al. / Clinical Biomechanics 30 (2015) 405–410
Fig. 5. Typical failure pattern. The specimens failed with a cortical break originating from the screw hole of the proximal fragment and proceeding towards the osteotomy. As indicated by the arrow, this leads to a loss of reduction.
Our findings suggest that there is no essential biomechanical advantage of bicortical over unicortical screw fixation in a phalangeal locking plate system, which is in line with the results of previous studies performed by other research groups (Dona et al., 2004; Freeland and Lindley, 2006; Freeland et al., 2003). Bicortical fixation is known to provide higher pullout strength compared to monocortical screw fixation (Khalid et al., 2008; Windolf and Perren, 2012), but does not necessarily imply a higher construct rigidity. Our study has shown that the predominant mode of failure is a fracture of the bone bridge between the screw and the osteotomy rather than screw pullout. Hence, this advantage of bicortical screw placement does not come into play in the investigated fracture model. Different results may be obtained using a different loading protocol as for example under torsional loading, which however might not be relevant for the early postoperative rehabilitation of plated phalanx (Dona et al., 2004). Therefore, this is an important statement since in clinical practice the use of unicortical screws minimizes the risk of irritation of the far cortex anatomic structures. Management of fractures around the hand is generally fraught with possible loss of function due to stiffness and reduced joint range of motion. Kurzen et al. (2006) stated that most fractures involving the phalanges would have some amount of functional loss and therefore require meticulous soft tissue care during any fracture treatment procedure in this anatomical region. In view of the limited space and close relationship of the tendons and neurovascular structures to the phalanx, the surgeon needs to be very conscious of the positioning of any implants. Lateral plating reduces the risk for complications caused by interference with the surrounding tendons (Lin et al., 2013). Lateral plating using the midaxial approach is safe as the neurovascular bundle lies in the palmar flap and thus is out of risk. In contrast, performing lateral plating from a dorsal approach may place those structures at risk if drilling and screw insertion is not done with care. On the other hand, the surgical approach and implant placement are more difficult compared to dorsal plate fixation. Potential problems can arise in case the plate is too bulky and causes irritation to the tendons or the soft tissue. The use of a smaller implant might reduce this risk. Our results implicate that the smaller 1.3 mm locking plate is biomechanically equivalent to the bigger conventional 1.5 mm non-locking plate. Considering the anatomy of the finger, the ideal place for plate positioning would be either on the lateral or the medial aspect of the bone. In this way, the plate would cause the least interference to the tendons which need to glide as smoothly and unimpeded as possible to achieve good function of the fingers. Furthermore, the risk of scarring of the tendon by its splitting, as well as formation of adhesions might be also reduced (Dabezies and Schutte, 1986; Kozin et al., 2000). This can be further enhanced by the use of unicortical screws since the far cortex does not need to be bored through, thus eliminating any chance of
409
injury to the structures beyond this cortex by the drill. However, this is only possible with the angular stability offered by locking plate systems (Marti et al., 2001; Szypryt and Forward, 2009) as non-locked unicortical screw fixation has been shown to be inferior to nonlocking bicortical screw fixation (Afshar et al., 2012). Non-locking plates work best when used on the tension side to provide absolute rigid fixation or as a neutralizing plate in addition to a lag screw. However, in the finger the tension and compression side may change depending on the patient action at any time, thus the ideal fixation device should offer sufficient stability in all relevant loading modes. A few limitations which are commonly associated with cadaveric studies should be acknowledged. There was a large scatter of results despite the BMD matching of the specimens. This might be due to the fact that human phalanges can vary considerably in size both within the same hand as well as between hands. To minimize this variation, we only used the 2nd to 4th phalanges but there are still some differences in specimen size that probably account for the rather large standard deviations observed in the study. Increasing specimen number and using phalanges from one single ray or performing paired tests might minimize this problem but is highly unpractical due to ethical reasons and limited availability of specimens. Although all specimens were prepared by a single surgeon, there might have been some variation in the compression of the fracture fragments which may have contributed to the scatter of the results. Furthermore, we used a transverse fracture model. For that reason, our results are not directly transferable to a clinical situation involving a comminuted fracture. Further biomechanical and especially clinical studies are necessary to investigate the advantage of lateral over dorsal plating with the use of the novel small 1.3 mm locking plate.
5. Conclusion Our study suggests that dorsal placement of a conventional 1.5 mm non-locking plate for proximal phalanx fracture has no significant biomechanical advantage over lateral placement of a locking plate. The novel low-profile 1.3 mm lateral locking plate provides significantly more rigid fixation compared to the conventional 1.5 mm non-locking plate in ML loading, and comparable fixation rigidity in DV loading direction using unicortical or bicortical screws. We therefore conclude that lateral plating of non-comminuted phalanx fractures using a smaller locking plate is a valid treatment option that might reduce the risk of soft tissue irritation and tendon injuries compared to conventional dorsal plating with a 1.5 mm non-locking plate. Contribution of authors R.S. and M.E. contributed equally to this study. R.S.: Study design, material and specimen instrumentation, writing. M.E.: Testing, data collection and processing, writing. K.S.: Literature research, analysis and interpretation, critical review. M.F.: Literature research, analysis and interpretation, writing. D.W.: Study design and test setup, technical assistance, data collection. G.R.: Concept, supervision, critical review. B.G.: Resources, Critical review, supervision, statistical evaluation. Acknowledgments The authors declare that they have no conflict of interest. This investigation was performed with the assistance of the AO Foundation via the AOTRAUMA Network (Grant No.: AR2012_01). The authors are not compensated and there are no other institutional subsidies, corporate affiliations, or funding sources supporting this work unless clearly documented and disclosed.
410
R. Shanmugam et al. / Clinical Biomechanics 30 (2015) 405–410
References Adams, J.E., Miller, T., Rizzo, M., 2013. The biomechanics of fixation techniques for hand fractures. Hand Clin. 29, 493–500. Afshar, R., Fong, T.S., Latifi, M.H., Kanthan, S.R., Kamarul, T., 2012. A biomechanical study comparing plate fixation using unicortical and bicortical screws in transverse metacarpal fracture models subjected to cyclic loading. J. Hand Surg. Eur. Vol. 37, 396–401. An, K.N., Chao, E.Y., Cooney, W.P., Linscheid, R.L., 1985. Forces in the normal and abnormal hand. J. Orthop. Res. 3, 202–211. Berman, K.S., Rothkopf, D.M., Shufflebarger, J.V., Silverman, R., 1999. Internal fixation of phalangeal fractures using titanium miniplates. Ann. Plast. Surg. 42, 408–410. Berme, N., Paul, J.P., Purves, W.K., 1977. A biomechanical analysis of the metacarpophalangeal joint. J. Biomech. 10, 409–412. Black, D.M., Mann, R.J., Constine, R.M., Daniels, A.U., 1986. The stability of internal fixation in the proximal phalanx. J. Hand. Surg. [Am.] 11, 672–677. Dabezies, E.J., Schutte, J.P., 1986. Fixation of metacarpal and phalangeal fractures with miniature plates and screws. J. Hand. Surg. [Am.] 11, 283–288. Dona, E., Gillies, R.M., Gianoutsos, M.P., Walsh, W.R., 2004. Plating of metacarpal fractures: unicortical or bicortical screws? J. Hand Surg. 29, 218–221. Egol, K.A., Kubiak, E.N., Fulkerson, E., Kummer, F.J., Koval, K.J., 2004. Biomechanics of locked plates and screws. J. Orthop. Trauma 18, 488–493. Fambrough, R.A., Green, D.P., 1979. Tendon rupture as a complication of screw fixation in fractures in the hand. A case report. J. Bone Joint Surg. (Am. Vol.) 61, 781–782. Firoozbakhsh, K.K., Moneim, M.S., Howey, T., Castaneda, E., Pirela-Cruz, M.A., 1993. Comparative fatigue strengths and stabilities of metacarpal internal fixation techniques. J. Hand Surg. 18, 1059–1068. Freeland, A.E., Lindley, S.G., 2006. Malunions of the finger metacarpals and phalanges. Hand Clin. 22, 341–355. Freeland, A.E., Hardy, M.A., Singletary, S., 2003. Rehabilitation for proximal phalangeal fractures. J. Hand Ther. 16, 129–142. Hardy, M.A., 2004. Principles of metacarpal and phalangeal fracture management: a review of rehabilitation concepts. J. Orthop. Sports Phys. Ther. 34, 781–799. Hattori, Y., Doi, K., Sakamoto, S., Yamasaki, H., Wahegaonkar, A., Addosooki, A., 2007. Volar plating for intra-articular fracture of the base of the proximal phalanx. J. Hand Surg. 32, 1299–1303. Haughton, D., Jordan, D., Malahias, M., Hindocha, S., Khan, W., 2012. Suppl 1: principles of hand fracture management. Open Orthop. J. 6, 43. Kamath, J.B., Harshvardhan, Naik, D.M., Bansal, A., 2011. Current concepts in managing fractures of metacarpal and phalangess. Indian J. Plast. Surg. 44, 203–211. Khalid, M., Theivendran, K., Cheema, M., Rajaratnam, V., Deshmukh, S.C., 2008. Biomechanical comparison of pull-out force of unicortical versus bicortical screws in proximal phalanges of the hand: a human cadaveric study. Clin. Biomech. 23, 1136–1140. Kozin, S.H., Thoder, J.J., Lieberman, G., 2000. Operative treatment of metacarpal and phalangeal shaft fractures. J. Am. Acad. Orthop. Surg. 8, 111–121.
Kurzen, P., Fusetti, C., Bonaccio, M., Nagy, L., 2006. Complications after plate fixation of phalangeal fractures. J. Trauma Acute Care Surg. 60, 841–843 (810.1097/ 1001.ta.0000214887.0000231745.c0000214884). Lin, S.Y., Huang, P.J., Huang, H.T., Chen, C.H., Cheng, Y.M., Fu, Y.C., 2013. An alternative technique for the management of phalangeal enchondromas with pathologic fractures. J. Hand. Surg. [Am.] 38, 104–109. Lu, W.W., Furumachi, K., Ip, W.Y., Chow, S.P., 1996. Fixation for comminuted phalangeal fractures. A biomechanical study of five methods. J. Hand Surg. 21, 765–767. Marti, A., Fankhauser, C., Frenk, A., Cordey, J., Gasser, B., 2001. Biomechanical evaluation of the less invasive stabilization system for the internal fixation of distal femur fractures. J. Orthop. Trauma 15, 482–487. Miller, D.L., Goswami, T., 2007. A review of locking compression plate biomechanics and their advantages as internal fixators in fracture healing. Clin. Biomech. 22, 1049–1062. Nunley, J.A., Kloen, P., 1991. Biomechanical and functional testing of plate fixation devices for proximal phalangeal fractures. J. Hand. Surg. [Am.] 16, 991–998. Ouellette, E.A., Dennis, J.J., Latta, L.L., Milne, E.L., Makowski, A.L., 2004. The role of soft tissues in plate fixation of proximal phalanx fractures. Clin. Orthop. Relat. Res. 213–218. Page, S.M., Stern, P.J., 1998. Complications and range of motion following plate fixation of metacarpal and phalangeal fractures. J. Hand Surg. 23, 827–832. Prevel, C.D., Eppley, B.L., Jackson, J.R., Moore, K., McCarty, M., Sood, R., Wood, R., 1995. Mini and micro plating of phalangeal and metacarpal fractures: a biomechanical study. J. Hand. Surg. [Am.] 20, 44–49. Pun, W., Chow, S., So, Y., Luk, K., Ngai, W., Ip, F., Peng, W., Ng, C., Crosby, C., 1991. Unstable phalangeal fractures: treatment by AO screw and plate fixation. J. Hand Surg. 16, 113–117. Ruchelsman, D.E., Mudgal, C.S., Jupiter, J.B., 2010. The role of locking technology in the hand. Hand Clin. 26, 307–319. Stern, P.J., Wieser, M.J., Reilly, D.G., 1987. Complications of plate fixation in the hand skeleton. Clin. Orthop. Relat. Res. 214, 59–65. Szypryt, P., Forward, D., 2009. The use and abuse of locking plates. Orthop. Trauma 23, 281–290. Windolf, M., Perren, S.M., 2012. Basic mechanobiology of bone healing. In: Babst, R., Bavonratanavech, S., Pesantez, R. (Eds.), Minimally Invasive Plate Osteosynthesis — Second, expanded edition AO Foundation/Georg Thieme, Stuttgart, New York, pp. 13–30. Wong, Hin-keung, Cho-yee, Lam, Wong, Kam-yiu, Ip, Wing-yuk, Fung, Kwok-keung, 2008. Treatment of phalangeal and metacarpal fractures: a review. Pb J. Orthop. X, 42–50. Wu, J.Z., An, K.N., Cutlip, R.G., Krajnak, K., Welcome, D., Dong, R.G., 2008. Analysis of musculoskeletal loading in an index finger during tapping. J. Biomech. 41, 668–676.
Contents lists available at ScienceDirect
Clinical Biomechanics journal homepage: www.elsevier.com/locate/clinbiomech
Angular stable lateral plating is a valid alternative to conventional plate fixation in the proximal phalanx. A biomechanical study R. Shanmugam a,b,1, M. Ernst a,⁎,1, K. Stoffel c,d, M.F. Fischer a, D. Wahl a, R.G. Richards a, B. Gueorguiev a a
AO Research Institute Davos, Davos, Switzerland Orthopaedic Department, University of Malaya, Kuala Lumpur, Malaysia Cantonal Hospital Baselland, Liestal, Switzerland d University of Basel, Basel, Switzerland b c
a r t i c l e
i n f o
Article history: Received 8 September 2014 Accepted 18 March 2015 Keywords: Proximal phalanx Lateral plating Fracture fixation
a b s t r a c t Background: Dorsal plating is commonly used in proximal phalanx fractures but it bears the risk of interfering with the extensor apparatus. In this study, dorsal and lateral plating fixation methods are compared to assess biomechanical differences using conventional 1.5 mm non-locking plates and novel 1.3 mm lateral locking plates. Methods: Twenty-four fresh frozen human cadaveric proximal phalanges were equally divided into four groups. An osteotomy was set at the proximal metaphyseal–diaphyseal junction and fixed with either dorsal (group A) or lateral (group B) plating using a 1.5 mm non-locking plate, or lateral plating with a novel 1.3 mm locking plate with bicortical (group C) or unicortical (group D) screws. The specimens were loaded in axial, dorsovolar and mediolateral direction to assess fixation stiffness followed by a cyclic destructive test in dorsovolar loading direction. Findings: Axial stiffness was highest in group D (mean 321.02, SEM 21.47 N/mm) with a significant difference between groups D and B (P = 0.033). Locking plates (groups C and D) were stiffer than non-locking plates under mediolateral loading (P = 0.007), no significant differences were noted under dorsovolar loading. Furthermore, no significant differences were observed under cyclic loading to failure between any of the study groups. Interpretation: No considerable biomechanical advantage of using a conventional 1.5 mm dorsal non-locking plate was identified over the novel 1.3 mm lateral locking plate in the treatment of proximal phalanx fractures. Since the novel low-profile plate is less disruptive to the extensor mechanism, it should be considered as a valid alternative. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Phalangeal fractures are among the commonest hand fractures constituting about 10% of all hand injuries (Kamath et al., 2011; Ouellette et al., 2004). Although a lot of phalangeal fractures can be managed non-operatively, operative treatment may be indicated in the presence of shortening, excessive deformity, involvement of multiple rays, open fractures or failure to achieve acceptable reduction (Adams et al., 2013; Wong et al., 2008; Kamath et al., 2011; Nunley and Kloen, 1991; Ouellette et al., 2004). Following hand fractures, early physiotherapy is important to regain a good functional outcome (Hardy, 2004; Lu et al., 1996; Nunley and Kloen, 1991; Page and Stern, 1998). This requires a method of fixation rigid enough to allow early commencement of physiotherapy without causing any loss of reduction or excessive movement
⁎ Corresponding author at: AO Research Institute Davos, Clavadelerstrasse 8, 7270 Davos Platz, Switzerland. E-mail address: [email protected] (M. Ernst). 1 These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.clinbiomech.2015.03.019 0268-0033/© 2015 Elsevier Ltd. All rights reserved.
at the fracture site that may be related to pain and/or non-unions (Freeland et al., 2003). Plating can hold an accurate fracture reduction rigidly and is suitable for fixation of phalangeal fractures as it would allow for early physiotherapy. There are two well described techniques for phalanx plating: the dorsal and the lateral technique (Berman et al., 1999; Hattori et al., 2007; Haughton et al., 2012; Ruchelsman et al., 2010). In dorsal plating, the incision is made over the dorsal aspect of the finger splitting the extensor tendon to access the bone. The plate is then placed on the dorsal aspect of the bone just underneath the extensor tendon. This often causes irritation during movement and might even lead to disruption of the tendon (Pun et al., 1991; Stern et al., 1987). The most frequent complication observed after open reduction and plate fixation of the phalanx is reduced mobility of the extensor tendons caused by scar tissue formation and resulting in a stiff joint (Kurzen et al., 2006; Page and Stern, 1998). In addition, in case of too long screws, their tips on the volar aspect might irritate the flexor tendon or even cause its rupture (Fambrough and Green, 1979; Kurzen et al., 2006). On the other hand, lateral plating carries a lower risk of irritation to the extensor mechanism as the respective tendon does not need to be split so that
406
R. Shanmugam et al. / Clinical Biomechanics 30 (2015) 405–410
scarring and adhesion are potentially avoided (Dabezies and Schutte, 1986; Kozin et al., 2000). Nevertheless, since the finger is a very compact structure, there is still some risk for the implant to be prominent or impinge on the neurovascular structures in close proximity or at the far cortex in case of bicortical screw fixation. Therefore, it is important to minimize surgical trauma during plating for good clinical outcomes applying a less invasive technique with low-profile implants in order to reduce possible irritation to surrounding tissues. Rigidity of various fixation methods was investigated in previous studies (Afshar et al., 2012; Black et al., 1986; Lu et al., 1996; Nunley and Kloen, 1991; Ouellette et al., 2004; Prevel et al., 1995). Most of them were conducted applying quasistatic loading, which is less representative for the repetitive movements encountered during rehabilitation, or three-point bending of the fingers with limited physiological relevance. In addition, there is no consensus regarding the predominant loading type of the fingers during rehabilitation following a phalanx fracture (An et al., 1985; Berme et al., 1977; Wu et al., 2008). In spite of this and the paucity of supporting data, dorsal plating is widely used for phalangeal fracture fixation due to its straightforward approach and the assumption that thereby a stabilizing tension band effect is achieved (Black et al., 1986; Prevel et al., 1995). One of the most commonly used implants for phalanx fractures is the T-plate (DePuy Synthes, Solothurn, Switzerland), available as 1.5 mm non-locking and locking plate, that can be used for both dorsal and lateral application by trimming and shaping the plate accordingly. Recently, a novel, anatomically shaped low-profile locking plate (DePuy Synthes, Solothurn, Switerzland, not yet available) has been designed especially for application on the lateral aspect of the proximal phalanx. The aim of this study was to compare the biomechanical performance of the novel low-profile lateral locking plate for proximal phalanx fracture fixation versus an established non-locking plate routinely used in clinics. We hypothesized that with the use of the locking screw technology a low-profile implant can reach comparable fixation rigidity as a bulkier conventional non-locking plate. Furthermore, it is expected that lateral plating is not biomechanically inferior to dorsal plating under consideration of various loading conditions. 2. Methods Twenty-four fresh frozen (−20 °C) human proximal phalanx cadaveric specimens from the 2nd to 4th fingers were used in this study. The specimens were thawed in a fridge (4 °C) for 24 h prior to preparation and biomechanical testing. Testing was performed at room temperature with phalanges kept moist. All specimens were subjected to bone mineral density (BMD) measurement using Xtreme CT (Scanco Medical, Brüttisellen, Switzerland) at a resolution of 82 μm. The region of interest was selected in the whole proximal phalanx from the proximal to the distal condyles encompassing the cortical and cancellous components. The mean computed value was considered as BMD of the particular phalanx. The 24 phalanges were equally divided into four study groups (Table 1) with no significant difference in the BMD distribution between the groups. Length of all specimens was set to 48 mm by cutting a part of the distal condyle in order to standardize the moment arm during the following cantilever bending tests. A transverse osteotomy was cut freehand 11 mm distally to the proximal joint line using an oscillating saw (Colibri system, DePuy Synthes, Solothurn, Switzerland) to simulate a
transverse proximal fracture at the metaphyseal–diaphyseal junction. The specimens were then subjected to fixation using two types of plates with different fixation methods as described in Table 1. All plates were trimmed to equalize the length and contoured to the bone as needed. In all study groups the proximal and the distal fragments were fixed with two screws each (Fig. 1). The fracture reduction was performed under slight compression and the fragments were fixed according to the manufacturer's guidelines (DePuy Synthes, Solothurn, Switzerland). Reproducibility of both fracture pattern and fixation technique with correct and consistent alignment and implants placement were ensured. K-wires were passed through both proximal and distal ends of each bone and incorporated into Polymethylmethacrylate (PMMA) (Troller AG, Fulenbach, Switzerland) to embed the specimens properly. Proximal and distal embedding was performed using custom made molds allowing the embedded specimens to fit into a jig during biomechanical testing (Fig. 2). The mold for embedding the distal end of the phalanx comprised a central opening at the bottom through which a metal pin was inserted before embedding. The specimen was then positioned with its distal end within the mold and the long bone axis aligned along the pin axis. The pin was incorporated in the PMMA block after embedding. All samples were prepared by a single surgeon (R.S.). Biomechanical testing was performed using a Bose 3220 material testing machine (BOSE, Eden Prairie, MN, USA) with a 225 N load cell coupled to a vertical slide with negligible friction. The specimens were mounted on a custom made jig allowing positioning in multiple directions according to the type of required loading (Fig. 3). For specimen's axial compression, the load was introduced via a beam connected to the vertical slide. For dorsovolar (DV) and mediolateral (ML) cantilever bending, the pin was inserted for loading into a hole in the vertical slide, keeping a 45-mm lever arm for each specimen. The load cell was coupled to the vertical slide. All specimens were firstly subjected to non-destructive axial compressive loading, followed by DV and ML cantilever bending to mimic early rehabilitation range-of-motion (ROM) exercises (Hardy, 2004; Nunley and Kloen, 1991). Finally, a destructive cyclic test in DV cantilever bending was run. Each of the three nondestructive tests comprised three successive quasistatic ramps at a rate of 0.01 mm/s. In case of axial compressive loading, each ramp ranged from 1 N (valley) to 5 N (peak) (Fig. 3A). During the nondestructive cantilever bending tests (Fig. 3B and C) the load was applied symmetrically at 0 N mean level and 3.0 N amplitude, except for the test phase when the specimen was loaded with the implant on the compression side. In this case, the force was limited to 1.5 N to avoid excessive opening of the osteotomy at the far cortex. The destructive sinusoidal cyclic test was performed at 0.5 Hz in DV loading direction until bone-implant construct failure. The valley load was kept at a constant level of 1 N throughout the whole test while the peak load, starting at 5 N, was progressively increased by 0.007 N every cycle. Based on the machine data acquired from the load-cell and system's transducer at a rate of 50 Hz, the following parameters of interest were evaluated. Stiffness of the bone-plate construct in all loading directions was calculated from the linear range of the respective loaddisplacement curve of the last of the three quasi-static nondestructive ramped cycles, the first two ramped cycles served as preconditioning of the specimen. Moreover, plastic deformation after 200 and 400 sinusoidal cycles was calculated as the relative axial
Table 1 Study groups. Details of the four study groups indicating the type of implant and fixation method applied.
Implant Plate position Screws
Group A (n = 6)
Group B (n = 6)
Group C (n = 6)
Group D (n = 6)
T-adaptation plate Dorsal 1.5 mm non-locking Bicortical
T-adaptation plate Lateral 1.5 mm non-locking Bicortical
Phalangeal base plate Lateral 1.3 mm locking Bicortical
Phalangeal base plate Lateral 1.3 mm locking Unicortical
R. Shanmugam et al. / Clinical Biomechanics 30 (2015) 405–410
Fig. 1. Instrumentation details. The length of the two implants used in this study was standardized by cutting off the last 5 (T-adaptation plate) or 2 holes (Phalangeal base plate), respectively. Additionally, the 1st of the three perpendicular holes in the lateral nonlocking group was cut for lateral plating (B). The two groups instrumented with the novel lateral locking plate (C and D) differ only in the use of bicortical (C) or unicortical (D) screws. Solid line indicates the section where the plate was cut; dotted line indicates the location of the osteotomy.
displacement at valley load of 1 N with respect to its value at valley load of the fifth cycle, taken as reference. Finally, cycles to failure were defined as the number of cycles until reaching a plastic deformation of 3 mm, considered as an arbitrary criterion for clinical failure. Statistical analysis was performed using SPSS (Version 19, IBM SPSS, Chicago, IL, USA). Normal distribution and homogeneity of variances of all parameters of interest in the study groups were tested with Shapiro– Wilk and Levene tests, respectively. Univariate Analysis of Variance (ANOVA) with Bonferroni Post Hoc tests, Paired-Samples t-test, as well as General Linear Model (GLM) Repeated Measures with a Greenhouse–Geisser epsilon adjustment were applied to detect significant differences, setting the level of significance at P = 0.05 for all statistical tests. 3. Results BMD was not significantly different between the study groups (A: mean 342.27, SEM (standard error of mean) 50.85 mgHA/ccm, B: mean 329.91, SEM 42.99 mgHA/ccm, C: mean 363.13, SEM 20.92 mgHA/ccm, D: mean 333.27, SEM 45.86 mgHA/ccm). Stiffness values, calculated for the various loading conditions in terms of mean and SEM, are shown in Fig. 4. Under axial loading, stiffness values were generally higher in the locking plate constructs in comparison to the non-locking ones. Statistical significant difference under axial loading was detected between group B (non-locked lateral plating with bicortical screws, mean 256.09, SEM 10.00 N/mm) and D (locked lateral plating with unicortical screws, mean 321.02, SEM 21.47 N/mm), P = 0.033. No significant differences between the groups were observed for DV stiffness (A: mean 8.22, SEM 2.14 N/mm, B: mean 6.36, SEM
407
1.43 N/mm, C: mean 9.22, SEM 2.77 N/mm, D: mean 5.85, SEM 1.83 N/mm), in group A evaluated during the phase of the DV cantilever bending test with the plate on tension side. However, under ML loading, the constructs with locking plates (C: mean 18.32, SEM 4.73 N/mm, D: mean 16.57, SEM 4.97 N/mm) were significantly stiffer compared to the non-locked constructs (A: mean 6.86, SEM 1.42 N/mm, with plate on tension side, B: mean 6.44, SEM 0.93 N/mm), P = 0.007. There were no significant differences in stiffness noted between the dorsal and lateral non-locking plate groups in all modes of loading as well as between unicortical and bicortical locked plate constructs. Stiffness evaluation during the compression and extension phase of each DV and ML cantilever bending test in each group separately revealed significantly higher values when the plate was on the tension side (compression phase) in comparison to its position on the compression specimen side (extension phase), P = 0.004. There was a difference between ML and DV stiffness in the locking plate groups (average difference 9.1 N/mm, with P = 0.021 and 10.7 N/mm, with P N 0.05 for groups C and D, respectively) with the specimens under ML loading being stiffer. In contrast, the bending stiffness did not significantly differ between the loading directions in the non-locked groups. In the dorsal non-locked group, bending stiffness was on average 1.36 N/mm higher under DV loading, and in the lateral non-locked group 0.08 N/mm higher under ML loading. Plastic deformation progressively increased throughout the cyclic destructive test in all groups. After 200 cycles, the highest plastic deformation was observed in the dorsal non-locking group with 0.71 mm on average (SEM 0.23 mm), followed by locked lateral plating with unicortical screws (mean 0.58 mm, SEM 0.16 mm), locked lateral plating with bicortical screws (mean 0.54 mm, SEM 0.16 mm) and the lowest plastic deformation in the lateral non-locking group with 0.43 mm on average (SEM 0.18 mm). Similarly, the values for plastic deformation after 400 cycles were highest in the dorsal non-locking group (mean 0.99 mm, SEM 0.39 mm), lowest in the non-locking lateral plating group (mean 0.55 mm, SEM 0.22 mm) and the two lateral locking plate groups in between (mean 0.82, SEM 0.27 mm with bicortical screws; mean 0.76 mm, SEM 0.27 mm with monocortical screws). The plastic deformation increased significantly in all groups between 200 and 400 cycles (P = 0.022), differences between the groups were not statistically significant. Specimens of the dorsal non-locked group withstood the highest number of cycles to failure (mean 2350, SEM 688 cycles), followed by the locked lateral group with bicortical screws (mean 1983, SEM 433), locked lateral plating with monocortical screws (mean 1618, SEM 352) and non-locked lateral plating (mean 1493, SEM 386). However, no statistical significance was screened with regard to this parameter of interest. Typical catastrophic failure observed in all specimens was breakage of the bone bridge from the screw hole of the closest screw on the proximal fragment towards the osteotomy (Fig. 5). This resulted in loss of reduction due to the screw displacing along the fracture, thereby often pushing a small bone fragment into the osteotomy. 4. Discussion
Fig. 2. Exemplified specimen after embedding in PMMA, ready to be mounted for testing.
Early rehabilitation of operatively treated phalanx fractures typically comprises ROM exercises as well as tendon gliding exercises (Hardy, 2004). These are generally repetitive in nature and are both active and passive but under low load. Hence, a cyclic loading regime would be ideal for testing. At the same time very little data is available on the actual loads on the phalanx during such exercises. Thus, a test protocol which gradually increases the loading with each cycle was chosen. This is in contrast with other studies that use quasistatic or fatigue type cyclic loading (Black et al., 1986; Firoozbakhsh et al., 1993; Prevel et al., 1995). When applying plating as a method of fracture fixation, it is essential to place the plate on the tension side of the fractured bone (Egol et al., 2004; Ouellette et al., 2004). This enhances stability of fixation through
408
R. Shanmugam et al. / Clinical Biomechanics 30 (2015) 405–410
Fig. 3. Illustration of the test setup and loading modes with given force ranges during biomechanical testing. A shows a specimen mounted for axial loading, B and C show a specimen prepared for ML and DV cantilever bending, respectively. The specimen is rotated along the long bone axis to change between ML and DV loading, while the test setup remains the same. Arrows indicate the direction of loading.
the far cortical support under compression, taking up a substantial amount of load rather than depending on the plate entirely. This is evidenced in the current study by the significantly higher bending stiffness achieved during the compression phase of loading (plate on the tension side) compared to its extension phase (plate on compression side). In addition, this also implies the presence of good fracture reduction in the current study resulting in load sharing between bone and plate on tension side for all specimens. Dorsal plating with a non-locking T-plate was tested as a control group in our study because this implant is established for surgical fixation of proximal phalanx fractures and to the authors knowledge provides sufficient construct stability for functional aftercare. None of the other three test groups performed significantly worse than the control group, therefore, it is concluded that all four tested constructs provide the required stability during the rehabilitation period. This statement is supported by the fact that early rehabilitation primarily concentrates
on ROM exercises to prevent joint stiffness, thereby avoiding excessive loading that could lead to failure of the bone-implant construct. In our study the locking implants revealed higher stiffness compared to the non-locking ones in axial and mediolateral loading directions. This difference was highly significant in mediolateral cantilever bending but less obvious under axial compression, where the good fracture reduction with optimal cortical support probably prevented larger differences. This finding is consistent with similar outcomes of previous studies (Egol et al., 2004; Miller and Goswami, 2007) and may be explained with the angular stability of the locked screw-plate connection which does not allow for any slipping or change in the angulation so that screw loosening and pullout can be prevented. Consequently, it ultimately leads to a stiffer construct and higher pullout resistance. For that reason the novel low-profile plates tested in the current study are stiffer than the larger 1.5 mm non-locking plates under most of the loading conditions.
Fig. 4. Stiffness results. Comparison of mean values for the stiffness evaluated under DV and ML bending (left) and under axial compression (right). Error bars indicate standard error of mean, * indicate significant differences (*P b 0.05, **P b 0.01).
R. Shanmugam et al. / Clinical Biomechanics 30 (2015) 405–410
Fig. 5. Typical failure pattern. The specimens failed with a cortical break originating from the screw hole of the proximal fragment and proceeding towards the osteotomy. As indicated by the arrow, this leads to a loss of reduction.
Our findings suggest that there is no essential biomechanical advantage of bicortical over unicortical screw fixation in a phalangeal locking plate system, which is in line with the results of previous studies performed by other research groups (Dona et al., 2004; Freeland and Lindley, 2006; Freeland et al., 2003). Bicortical fixation is known to provide higher pullout strength compared to monocortical screw fixation (Khalid et al., 2008; Windolf and Perren, 2012), but does not necessarily imply a higher construct rigidity. Our study has shown that the predominant mode of failure is a fracture of the bone bridge between the screw and the osteotomy rather than screw pullout. Hence, this advantage of bicortical screw placement does not come into play in the investigated fracture model. Different results may be obtained using a different loading protocol as for example under torsional loading, which however might not be relevant for the early postoperative rehabilitation of plated phalanx (Dona et al., 2004). Therefore, this is an important statement since in clinical practice the use of unicortical screws minimizes the risk of irritation of the far cortex anatomic structures. Management of fractures around the hand is generally fraught with possible loss of function due to stiffness and reduced joint range of motion. Kurzen et al. (2006) stated that most fractures involving the phalanges would have some amount of functional loss and therefore require meticulous soft tissue care during any fracture treatment procedure in this anatomical region. In view of the limited space and close relationship of the tendons and neurovascular structures to the phalanx, the surgeon needs to be very conscious of the positioning of any implants. Lateral plating reduces the risk for complications caused by interference with the surrounding tendons (Lin et al., 2013). Lateral plating using the midaxial approach is safe as the neurovascular bundle lies in the palmar flap and thus is out of risk. In contrast, performing lateral plating from a dorsal approach may place those structures at risk if drilling and screw insertion is not done with care. On the other hand, the surgical approach and implant placement are more difficult compared to dorsal plate fixation. Potential problems can arise in case the plate is too bulky and causes irritation to the tendons or the soft tissue. The use of a smaller implant might reduce this risk. Our results implicate that the smaller 1.3 mm locking plate is biomechanically equivalent to the bigger conventional 1.5 mm non-locking plate. Considering the anatomy of the finger, the ideal place for plate positioning would be either on the lateral or the medial aspect of the bone. In this way, the plate would cause the least interference to the tendons which need to glide as smoothly and unimpeded as possible to achieve good function of the fingers. Furthermore, the risk of scarring of the tendon by its splitting, as well as formation of adhesions might be also reduced (Dabezies and Schutte, 1986; Kozin et al., 2000). This can be further enhanced by the use of unicortical screws since the far cortex does not need to be bored through, thus eliminating any chance of
409
injury to the structures beyond this cortex by the drill. However, this is only possible with the angular stability offered by locking plate systems (Marti et al., 2001; Szypryt and Forward, 2009) as non-locked unicortical screw fixation has been shown to be inferior to nonlocking bicortical screw fixation (Afshar et al., 2012). Non-locking plates work best when used on the tension side to provide absolute rigid fixation or as a neutralizing plate in addition to a lag screw. However, in the finger the tension and compression side may change depending on the patient action at any time, thus the ideal fixation device should offer sufficient stability in all relevant loading modes. A few limitations which are commonly associated with cadaveric studies should be acknowledged. There was a large scatter of results despite the BMD matching of the specimens. This might be due to the fact that human phalanges can vary considerably in size both within the same hand as well as between hands. To minimize this variation, we only used the 2nd to 4th phalanges but there are still some differences in specimen size that probably account for the rather large standard deviations observed in the study. Increasing specimen number and using phalanges from one single ray or performing paired tests might minimize this problem but is highly unpractical due to ethical reasons and limited availability of specimens. Although all specimens were prepared by a single surgeon, there might have been some variation in the compression of the fracture fragments which may have contributed to the scatter of the results. Furthermore, we used a transverse fracture model. For that reason, our results are not directly transferable to a clinical situation involving a comminuted fracture. Further biomechanical and especially clinical studies are necessary to investigate the advantage of lateral over dorsal plating with the use of the novel small 1.3 mm locking plate.
5. Conclusion Our study suggests that dorsal placement of a conventional 1.5 mm non-locking plate for proximal phalanx fracture has no significant biomechanical advantage over lateral placement of a locking plate. The novel low-profile 1.3 mm lateral locking plate provides significantly more rigid fixation compared to the conventional 1.5 mm non-locking plate in ML loading, and comparable fixation rigidity in DV loading direction using unicortical or bicortical screws. We therefore conclude that lateral plating of non-comminuted phalanx fractures using a smaller locking plate is a valid treatment option that might reduce the risk of soft tissue irritation and tendon injuries compared to conventional dorsal plating with a 1.5 mm non-locking plate. Contribution of authors R.S. and M.E. contributed equally to this study. R.S.: Study design, material and specimen instrumentation, writing. M.E.: Testing, data collection and processing, writing. K.S.: Literature research, analysis and interpretation, critical review. M.F.: Literature research, analysis and interpretation, writing. D.W.: Study design and test setup, technical assistance, data collection. G.R.: Concept, supervision, critical review. B.G.: Resources, Critical review, supervision, statistical evaluation. Acknowledgments The authors declare that they have no conflict of interest. This investigation was performed with the assistance of the AO Foundation via the AOTRAUMA Network (Grant No.: AR2012_01). The authors are not compensated and there are no other institutional subsidies, corporate affiliations, or funding sources supporting this work unless clearly documented and disclosed.
410
R. Shanmugam et al. / Clinical Biomechanics 30 (2015) 405–410
References Adams, J.E., Miller, T., Rizzo, M., 2013. The biomechanics of fixation techniques for hand fractures. Hand Clin. 29, 493–500. Afshar, R., Fong, T.S., Latifi, M.H., Kanthan, S.R., Kamarul, T., 2012. A biomechanical study comparing plate fixation using unicortical and bicortical screws in transverse metacarpal fracture models subjected to cyclic loading. J. Hand Surg. Eur. Vol. 37, 396–401. An, K.N., Chao, E.Y., Cooney, W.P., Linscheid, R.L., 1985. Forces in the normal and abnormal hand. J. Orthop. Res. 3, 202–211. Berman, K.S., Rothkopf, D.M., Shufflebarger, J.V., Silverman, R., 1999. Internal fixation of phalangeal fractures using titanium miniplates. Ann. Plast. Surg. 42, 408–410. Berme, N., Paul, J.P., Purves, W.K., 1977. A biomechanical analysis of the metacarpophalangeal joint. J. Biomech. 10, 409–412. Black, D.M., Mann, R.J., Constine, R.M., Daniels, A.U., 1986. The stability of internal fixation in the proximal phalanx. J. Hand. Surg. [Am.] 11, 672–677. Dabezies, E.J., Schutte, J.P., 1986. Fixation of metacarpal and phalangeal fractures with miniature plates and screws. J. Hand. Surg. [Am.] 11, 283–288. Dona, E., Gillies, R.M., Gianoutsos, M.P., Walsh, W.R., 2004. Plating of metacarpal fractures: unicortical or bicortical screws? J. Hand Surg. 29, 218–221. Egol, K.A., Kubiak, E.N., Fulkerson, E., Kummer, F.J., Koval, K.J., 2004. Biomechanics of locked plates and screws. J. Orthop. Trauma 18, 488–493. Fambrough, R.A., Green, D.P., 1979. Tendon rupture as a complication of screw fixation in fractures in the hand. A case report. J. Bone Joint Surg. (Am. Vol.) 61, 781–782. Firoozbakhsh, K.K., Moneim, M.S., Howey, T., Castaneda, E., Pirela-Cruz, M.A., 1993. Comparative fatigue strengths and stabilities of metacarpal internal fixation techniques. J. Hand Surg. 18, 1059–1068. Freeland, A.E., Lindley, S.G., 2006. Malunions of the finger metacarpals and phalanges. Hand Clin. 22, 341–355. Freeland, A.E., Hardy, M.A., Singletary, S., 2003. Rehabilitation for proximal phalangeal fractures. J. Hand Ther. 16, 129–142. Hardy, M.A., 2004. Principles of metacarpal and phalangeal fracture management: a review of rehabilitation concepts. J. Orthop. Sports Phys. Ther. 34, 781–799. Hattori, Y., Doi, K., Sakamoto, S., Yamasaki, H., Wahegaonkar, A., Addosooki, A., 2007. Volar plating for intra-articular fracture of the base of the proximal phalanx. J. Hand Surg. 32, 1299–1303. Haughton, D., Jordan, D., Malahias, M., Hindocha, S., Khan, W., 2012. Suppl 1: principles of hand fracture management. Open Orthop. J. 6, 43. Kamath, J.B., Harshvardhan, Naik, D.M., Bansal, A., 2011. Current concepts in managing fractures of metacarpal and phalangess. Indian J. Plast. Surg. 44, 203–211. Khalid, M., Theivendran, K., Cheema, M., Rajaratnam, V., Deshmukh, S.C., 2008. Biomechanical comparison of pull-out force of unicortical versus bicortical screws in proximal phalanges of the hand: a human cadaveric study. Clin. Biomech. 23, 1136–1140. Kozin, S.H., Thoder, J.J., Lieberman, G., 2000. Operative treatment of metacarpal and phalangeal shaft fractures. J. Am. Acad. Orthop. Surg. 8, 111–121.
Kurzen, P., Fusetti, C., Bonaccio, M., Nagy, L., 2006. Complications after plate fixation of phalangeal fractures. J. Trauma Acute Care Surg. 60, 841–843 (810.1097/ 1001.ta.0000214887.0000231745.c0000214884). Lin, S.Y., Huang, P.J., Huang, H.T., Chen, C.H., Cheng, Y.M., Fu, Y.C., 2013. An alternative technique for the management of phalangeal enchondromas with pathologic fractures. J. Hand. Surg. [Am.] 38, 104–109. Lu, W.W., Furumachi, K., Ip, W.Y., Chow, S.P., 1996. Fixation for comminuted phalangeal fractures. A biomechanical study of five methods. J. Hand Surg. 21, 765–767. Marti, A., Fankhauser, C., Frenk, A., Cordey, J., Gasser, B., 2001. Biomechanical evaluation of the less invasive stabilization system for the internal fixation of distal femur fractures. J. Orthop. Trauma 15, 482–487. Miller, D.L., Goswami, T., 2007. A review of locking compression plate biomechanics and their advantages as internal fixators in fracture healing. Clin. Biomech. 22, 1049–1062. Nunley, J.A., Kloen, P., 1991. Biomechanical and functional testing of plate fixation devices for proximal phalangeal fractures. J. Hand. Surg. [Am.] 16, 991–998. Ouellette, E.A., Dennis, J.J., Latta, L.L., Milne, E.L., Makowski, A.L., 2004. The role of soft tissues in plate fixation of proximal phalanx fractures. Clin. Orthop. Relat. Res. 213–218. Page, S.M., Stern, P.J., 1998. Complications and range of motion following plate fixation of metacarpal and phalangeal fractures. J. Hand Surg. 23, 827–832. Prevel, C.D., Eppley, B.L., Jackson, J.R., Moore, K., McCarty, M., Sood, R., Wood, R., 1995. Mini and micro plating of phalangeal and metacarpal fractures: a biomechanical study. J. Hand. Surg. [Am.] 20, 44–49. Pun, W., Chow, S., So, Y., Luk, K., Ngai, W., Ip, F., Peng, W., Ng, C., Crosby, C., 1991. Unstable phalangeal fractures: treatment by AO screw and plate fixation. J. Hand Surg. 16, 113–117. Ruchelsman, D.E., Mudgal, C.S., Jupiter, J.B., 2010. The role of locking technology in the hand. Hand Clin. 26, 307–319. Stern, P.J., Wieser, M.J., Reilly, D.G., 1987. Complications of plate fixation in the hand skeleton. Clin. Orthop. Relat. Res. 214, 59–65. Szypryt, P., Forward, D., 2009. The use and abuse of locking plates. Orthop. Trauma 23, 281–290. Windolf, M., Perren, S.M., 2012. Basic mechanobiology of bone healing. In: Babst, R., Bavonratanavech, S., Pesantez, R. (Eds.), Minimally Invasive Plate Osteosynthesis — Second, expanded edition AO Foundation/Georg Thieme, Stuttgart, New York, pp. 13–30. Wong, Hin-keung, Cho-yee, Lam, Wong, Kam-yiu, Ip, Wing-yuk, Fung, Kwok-keung, 2008. Treatment of phalangeal and metacarpal fractures: a review. Pb J. Orthop. X, 42–50. Wu, J.Z., An, K.N., Cutlip, R.G., Krajnak, K., Welcome, D., Dong, R.G., 2008. Analysis of musculoskeletal loading in an index finger during tapping. J. Biomech. 41, 668–676.
