- Email: [email protected]
Plate fixation of comminuted olecranon fractures: An in vitro biomechanical study
Plate fixation of comminuted olecranon fractures: An in vitro biomechanical study G. J. W. King, MD, P. N. Lammens, MD, A. D. Milne, BEng, J. H. Roth, MD, and J. A. Johnson, PhD, London, Ontario, Canada
The stiffness, cyclic stability, and failure strength of posterior and lateral plating were compared in an in vitro model of a comminuted olecranon fracture. Loading was applied to the brachialis and triceps while displacements of the olecranon were measured with an electromagnetic tracking device in six degrees of freedom. No statistical difference was seen in the cyclic or static stability of either plating method. The triceps tendon suture attachment failed during destructive testing in each case between 300 and 500 N of applied loading. No gross failure of the bony fixation of the implant occurred before suture failure. These results suggest that both plating methods are likely to afford adequate stability to permit early protected postoperative range of motion. (J SHOULDERELBowSuRe 1996;5:437-4 1.) Tension band wiring of comminuted olecranon fractures has been complicated by a high incidence of fixation failure and hardware migration.4, s, z, ~, lO, 13 Recent clinical and biomechanical studies suggest that plate fixation of displaced olecranon fractures yields improved results.3"~' 7, 12 The clinical use of posteriorly placed implants on the olecranon has been problematic because of poor soft-tissue coverage, resulting in wound breakdown and hardware discomfort necessitating plate removal. Clinical exp.erience using a lateral plate for fractures of the olecranon has demonstrated that this method is technically possible and is well tolerated by patients. The plate is placed beneath the anconeus muscle and wraps around the olecranon deep to the triceps tendon. Bicortical fixation is achieved in both the olecranon and the ulnar shaft. Should the stability afforded by the plate in this location be adequate, the clinical results of olecranon plating may be improved because of greater coverage of the hardware by soft tissue. The purpose of this investigation was to From the Musculoskeletal Research Laboratory, Hand & Upper Limb Centre, Department of Surgery, University of Western Ontario. Reprint requests: Graham J. W. King, MD, Hand & Upper Limb Centre, St. Joseph's Health Centre, 268 GrosvenorSt., London, Ontario N6A 4V2 Canada. Copyright @ 1996 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-27456/96/$5.00 + 0 3211176559
compare the stiffness and strength of posterior with lateral plating in simulated comminuted olecranon fractures. METHODS
Eight fresh, previously frozen, cadaveric elbows were used in this investigation. The humeral shaft was cleared of soft tissue and mounted on a custom Delrin clamp with the elbow unconstrained except in flexion and extension as shown in Figure 1. Stainless steel cables were sutured to the brachialis and triceps tendons and loaded by a system of pulleys attached to a materials testing machine (Instron Corporation, Canton, Mass.). This arrangement ensured that the loads applied to the brachialis and triceps tendons were equivalent throughout both static and cyclic testing. Care was taken to replicate the normal muscle moment arms of these tendons.1 Six-hole 3.5 mm pelvic reconstruction plates (Synthes Canada Ltd., Mississauga, Ontario, Canada) were contoured to both the lateral and posterior aspects of the ulnae. Trial plate application was performed, and all screws were predrilled and tapped to ensure an anatomic reduction after the olecranon osteotomy. Three fully threaded cancellous screws were used in the proximal olecranon fragment, and three cortical screws were used in the ulna for each plate location. After contouring was performed, the plates were removed, and a transverse segment of ulna measuring 3 mm was resected from the nonarticular mid437
438
King etal.
D
J. Shoulder Elbow Surg. November/December 1996
Q
\
Figure 1 Experimentalsetup. Humerus is fixed to Dekin testing jig, ancl cables are secured to brachialis and triceps with heavy suture. Electromagnetic tracking sensors are secured proximal and distal to osteotomy to measure differential motion be~een fragments. portion of the greater sigmoid notch to simulate a comminuted olecranon fracture. By excising a segment of olecranon, no contact occurred between the two fragments, making the model a test of rigidity of the plate/screw construct. The fracture site was not compressed by the fixation, because this would have distorted the greater sigmoid notch, similar to what occurs clinically with comminuted olecranon fractures. The experimental design block randomized the location in which the plate was applied initially for each specimen (Figure 2). Static testing assessed the fixation rigidity of the olecranon by applying equal forces concurrently to the triceps and brachialis tendons from an initial preload of 12.5 N to a maximum of 100 N. This loading range was selected to simulate that which may occur clinically in the postoperative period with early unresisted passive motion. This static stiffness test was performed at 30 ~, 60 ~, 90 ~, and 120 ~ of elbow flexion. After static testing was performed, cyclic loading for 4000 cycles at a frequency of 1 Hz was applied with the elbow at 90 ~ of flexion. This consisted of sinusoidal loading from 25 to 100 N over a period of 67 minutes. The plate location and screws were then changed, and the static and cyclic testing protocol was repeated with the plate and screws in the second position. The ultimate strength of the olecranon fixation was then tested with the elbow at 90 ~ of flexion and incremental loading of the triceps tendon until failure of the fixation or rupture of the triceps
Figure 2 Plating methods. A, Radiograph of lateral plate across decranon osteotomy. B, Radiograph of posterior plate across olecranon osteotomy. tendon suture occurred. The failure test was performed for only one plate location in each specimen, with loading confined to the triceps tendon. A Flock of Birds (Ascension Technology Corporation, Colchester, Vt.) electromagnetic tracking device was used to determine the relative motion of the proximal ulnar fragment to the distal ulna. This device consists of a source that emits an electromagnetic field and sensors that monitor this magnetic field and determine their position and orientation relative to the source. This system provides six degree of freedom kinematic data at 100 Hz. We have determined the mean error of this device to be less than 2% of the measured rotations and translations with a resolution of 0.1 o and 0.25 mm, respectively. ~1 All materials in the field of measurement have previously been demonstrated to have no effect on the accuracy of the tracking system. 11 One sensor was placed on the
J. Shoulder Elbow Surg. Volume 5, Number 6
King etal.
439
Table I Static stiffness Translation (mm)
Rotation (degrees) Angle (~ 30 60 90 120
Plate Poser• ~ral Poser• Lateral Posterior Lateral Pos~rior Lateral
Flexionextension 0.4• 0.8• 0.5• 1.2• 0.7• 0.8• 0.5• 0.8•
Valgusvarus
Internalexternal
Anteriorposterior
Mediallateral
0.2• 0.6• 0.5• 0.4• 0.8• 0.4• 0.6• 0.5•
0.4• 0.8• 0.6• 1.3• 1.2• 1.1• 1.1• 1.4•
0.6• 1.1• 0.5• 1.1• 0.6• 1.0• 0.7• 1.4•
0.3• O.7• 0.4• 0.4• 1.0• 0.4• 0.6• 0.7•
olecranon and one on the distal fragment of the ulna while the transmitter (source) was securely fixed to the testing table. The differential motion between the two fragments was calculated from the data generated by the electromagnetic tracking device. This allowed the determination of implant stiffness in six degrees of freedom. A stylus attached to a third sensor was used to digitize appropriate anatomic markers on the ulna from which a coordinate system was defined. This consisted of digitizing the center of the proximal fracture surface (Point O), the anterior apex of the proximal fracture surface, and the distal center of the ulnar shaft to establish the proximal-distal reference axis. The cross-product of these two vectors (i.e., with a common origin at the center) predetermined the medial-lateral reference axis. Subsequently, the cross-product of this axis with the proximal-distal reference axis established the anteriot-posterior reference axis. The original data from the tracking system was reduced to measure varus-valgus, flexion-extension, and internal-external rotations. The reference frames established for the olecranon and distal ulna were initially superimposed at point O before loading was performed. With the Euler sequence of rotations about the anterior-posterior, medial-lateral and proximaldistal axes of the olecranon, the relative rotation of the reference frame on the olecranon to the distal ulna was determined from basic kinematic algebra of the rotation matrixes. Translations relative to the fracture site were determined by calculating displacement of the olecranon coordinate frame origin from the distal ulnar frame, resolved along the anterior-posterior, medial-lateral, and proximal-distal axes. The static stiffness of the fixation was calculated
Proximaldistal 0.6• 0.7• 0.4• 0.7• 0.7• 0.6• 0.6• 0.6•
as the difference in the relative position of the olecranon to the distal ulna between 12.5 N and 100 N of applied loading. Behavior under cyclic loading was calculated as the difference in relative motion between the initial and final cycle of load application. The block randomized design allowed a comparison of the stiffness and fatigue behaviour of the two methods of plate fixation within the same specimens. The static stiffness data were analyzed with a two-way repeated measures analysis of variance with fixation method and flexion angles as the factors. Plate behavior under cyclic loading was analyzed with a repeated measures one-way analysis of variance. Failure strength was compared with attest. All statistical comparisons were conducted independently for the three rotations and the three translational motions measured, with 0~set at 0.05. RESULTS
Static stiffness testing yielded similar stability at all angles of elbow flexion (p > 0.05) (Table I). No significant difference was seen in the rotations or translations of the proximal olecranon fragment relative to the distal ulna between the two plating methods (p > 0.05). All of the motions were small and near the resolution of the tracking device. Cyclic loading of the specimens resulted in minimal changes in stiffness of the plate constructs over time (Figure 3). No statistical difference was seen in stability between the posterior or lateral plating subjected to cyclic loading (p > 0.05). Rupture of the triceps tendon suture attachment routinely occurred during failure testing at 300 to 500 N of applied force. No gross failure of the hardware occurred in any of the specimens during any phase of the experimental testing protocol.
440
J. Shoulder Elbow Surg. November/December 1996
King et al.
0.5
A
~"
0.4
~
0.3
~ '
0.2
~
0.1
~
0.0
!! F-E
V-V
0.4
--
A
E E
"-" Z iii LLI 0
"
D. (/3 El I-E
030.2-
0.1 0.0---
B
[--7
M-L
A-P
P-D
POSTERIOR LATERAL
Figure 3 Cyclic loading. Magnitude of cyclic stiffness of constructs with elbow held at 90 ~ of flexion for both rotation (A) and translation (B). There were no statistical differences be~een plating methods after 4000 cycles of loading (p > 0.05). FE, Flexion-extension; VV, valgus-varus; IE, internal-external rotation; ML, medial-lateral; AP, anterior-posterior; PD, proximal-distal. Our results demonstrate no significant difference in stability between posterior and lateral plating up to 300 N of triceps tendon loading (p > 0.05). DISCUSSION In this in vitro study we attempted to determine whether posterior and lateral plating of olecranon fractures afforded adequate stability to initiate early protected range of motion. The results of this investigation demonstrate that no significant differences exist in the low load stability of posterior and lateral plating. All of the measured displacements were small and probably not clinically significant. The measured rotations and translations were near the resolution of the tracking device (angulation 0.1 o and displacement 0.25 mm), 11 and therefore smaller differences in stability would not have been detected with this testing protocol. We chose a simulated comminuted fracture model as a "worse case scenario" of fixation stability and measured unconstrained translational and rotational motion at the osteotomy site in six degrees of freedom. The use of elderly cadaveric specimens with poor bone quality suggests that the clinical application of these techniques in younger patients should have even greater strength than reported in this study. Pelvic reconstruction plates were chosen because they are easy to contour to the shape of the olecranon. Although 3.5 dynamic
compression plates have greater rigidity than the pelvic reconstruction plates, they are difficult to contour and are more prominent. Dynamic compression plates may be useful when the olecranon fracture extends into the ulnar diaphysis; however, they are probably not necessary for isolated metaphyseal fractures where the fixation of the screws to bone is probably the weak link. A power analysis was performed to determine the probability of finding a difference in the stability of the two plating methods, if such a difference was present. The minimal clinically significant difference in stability to detect was set at 1 mm and 1~ For a sample size of eight specimens 13 was greater than 0.80 for all measured parameters in both the cyclic and static loading tests. This result suggests that the sample size of the study was sufficient to detect important clinical differences in stiffness between the two plating methods. We selected a level of loading of the brachialis and triceps that would be in the range anticipated for early unresisted active motion during rehabilitation. The tension generated in these tendons with both restricted and normal activity is unknown. Based on published data of muscle moment arms, loading of the triceps tendon to 100 N at a moment arm of 20 mm would balance an anteriorly directed force of approximately 7 N at the wrist (at a length of 300 mm)." 2. 9
King etal.
J. Shoulder Elbow Surg. Volume 5, Number 6
An interesting finding in this study was the somewhat complex and inconsistent motion patterns of the olecranon that emerged for "unidirectional" loading of the triceps tendon. The advantage of an electromagnetic tracking system over that of conventional testing of fixation devices is that it allows unconstrained testing with physiologic loading and assessment of fragment motion in six degrees of freedom. Out-of-plane motions may be an important cause of fixation failure that may go undetected with the more constraining conventional three- or four-point bending tests on a materials testing machine. The ultimate failure loads of the two fixation methods could not be measured because of the failure of the triceps tendon-suture attachment. Whether the absolute failure strength is important in plate fixation is debatable, because clinical failures are usually not catastrophic but rather tend to be due to repetitive loading across the fracture site. We chose cyclic loading as our major factor of interest in an attempt to mimic loading that may occur clinically with early postoperative elbow motion. In certain fracture patterns posterior plating of the olecranon will still be indicated. Patients with displaced coronoid fractures requiring internal fixation should probably have the plate applied posteriorly to allow the coronoid to be lagged back to the plate at the time of the internal fixation. Posterior comminution or extremely small olecranon fragments may also be an indication for posterior plating. Most fractures of the olecranon, however, do not have an associated coronoid fracture, and therefore lateral plating may be used. This method may be particularly indicated for simple olecranon fractures, where a reliable method of fixation is desired with improved softtissue coverage of the hardware. This method may result in a lower incidence of wound complications and postoperative hardware pain than has been reported with tension band wiring or posterior plating. Because of the lateral location of the plate, bicortical fixation of the proximal olecranon fragment may be possible with lateral plating, poten-
441
tially increasing the fixation strength. In this in vitro study both methods of plate fixation afforded adequate stability to permit early protected motion. Before this method can be applied clinically, however, a prospective clinical trial comparing the two plate fixation methods will be required. The authors appreciate the support of Synthes Canada in providing the internal fixation devices used in this study.
REFERENCES
1. An KN, Hui FC, Morrey BF, LinscheidRL, Chao EY. Muscles across the elbow joint: a biomechanical analysis. J Biomechanics 1981 ; 14:659-69. 2. Funk DA, An KN, Morrey BF, DaubeJR. Ebclromyographic analysis of muscles across the elbow joint, J Orthop Res 1987;5:529-38. 3. Fyfe IS, Mossad MM, Holdsworth BJ. Methods of fixation of olecranon fractures: an experimental mechanical study. J BoneJoint Surg Br 1985;67B:367-72. 4. Gartsman GM, Sculco TP, Otis JC. Operative lreatment of decranon fractures. J BoneJoint Surg Am 1981 ;63A;71821. 5. Home JG, TanzerTL. Olecranon fractures: a review of 100 Cases. J Trauma 1981 ;21:469-72. 6. Homer SR, Sadasivan KK, Lipka JM, Saha S. Analysis of mechanical tactors affecting fixation of olecranon fractures. Orthopaedics 1989;12:1469-72. 7. Hume MC, Wiss DA. Olecranon fractures: a clinical and radiographic comparison of tension band wiring and plate fixation. Clin Orthop 1992;285:229-35. 8. Johnson RP, Roetker A, Schwab JF. Olecranon fractures treated with AO screw and lension bands. Orthopaedics 1986;9:66-8. 9. King GJW, Itoi E, Niebur GL, Morrey BF, An KN. Motion and laxily of the capitdlocondylar total elbow arthroplasly. J BoneJoint Surg Am 1994;76A:1000-8. 10. Macko D, Szabo RM. Complications of tension-bandwiring of olecranon fractures. J Bone Joint Surg Am 1985;67A: 1396401. 11. Milne AD, JohnsonJA, ChessDG, King GJW. Accuracyof an electromagnetic tracking device: a study of the optimal operating range and metal interference. J Biomechanics ] 996;29:791-3. 12. Murphy DF, Greene WB, GilbertJA, DameronTB. Displaced olecranon fractures in adults: biomechanical analysis of fixation methods. Clin Orthop 1987;224:210-4. 13. Murphy DF, Greene WB, GilbertJA, DameronTB. Displaced olecranon fracturesin adults: clinical evaluation; Clin Orthop 1987;224:215-23.
The stiffness, cyclic stability, and failure strength of posterior and lateral plating were compared in an in vitro model of a comminuted olecranon fracture. Loading was applied to the brachialis and triceps while displacements of the olecranon were measured with an electromagnetic tracking device in six degrees of freedom. No statistical difference was seen in the cyclic or static stability of either plating method. The triceps tendon suture attachment failed during destructive testing in each case between 300 and 500 N of applied loading. No gross failure of the bony fixation of the implant occurred before suture failure. These results suggest that both plating methods are likely to afford adequate stability to permit early protected postoperative range of motion. (J SHOULDERELBowSuRe 1996;5:437-4 1.) Tension band wiring of comminuted olecranon fractures has been complicated by a high incidence of fixation failure and hardware migration.4, s, z, ~, lO, 13 Recent clinical and biomechanical studies suggest that plate fixation of displaced olecranon fractures yields improved results.3"~' 7, 12 The clinical use of posteriorly placed implants on the olecranon has been problematic because of poor soft-tissue coverage, resulting in wound breakdown and hardware discomfort necessitating plate removal. Clinical exp.erience using a lateral plate for fractures of the olecranon has demonstrated that this method is technically possible and is well tolerated by patients. The plate is placed beneath the anconeus muscle and wraps around the olecranon deep to the triceps tendon. Bicortical fixation is achieved in both the olecranon and the ulnar shaft. Should the stability afforded by the plate in this location be adequate, the clinical results of olecranon plating may be improved because of greater coverage of the hardware by soft tissue. The purpose of this investigation was to From the Musculoskeletal Research Laboratory, Hand & Upper Limb Centre, Department of Surgery, University of Western Ontario. Reprint requests: Graham J. W. King, MD, Hand & Upper Limb Centre, St. Joseph's Health Centre, 268 GrosvenorSt., London, Ontario N6A 4V2 Canada. Copyright @ 1996 by Journal of Shoulder and Elbow Surgery Board of Trustees. 1058-27456/96/$5.00 + 0 3211176559
compare the stiffness and strength of posterior with lateral plating in simulated comminuted olecranon fractures. METHODS
Eight fresh, previously frozen, cadaveric elbows were used in this investigation. The humeral shaft was cleared of soft tissue and mounted on a custom Delrin clamp with the elbow unconstrained except in flexion and extension as shown in Figure 1. Stainless steel cables were sutured to the brachialis and triceps tendons and loaded by a system of pulleys attached to a materials testing machine (Instron Corporation, Canton, Mass.). This arrangement ensured that the loads applied to the brachialis and triceps tendons were equivalent throughout both static and cyclic testing. Care was taken to replicate the normal muscle moment arms of these tendons.1 Six-hole 3.5 mm pelvic reconstruction plates (Synthes Canada Ltd., Mississauga, Ontario, Canada) were contoured to both the lateral and posterior aspects of the ulnae. Trial plate application was performed, and all screws were predrilled and tapped to ensure an anatomic reduction after the olecranon osteotomy. Three fully threaded cancellous screws were used in the proximal olecranon fragment, and three cortical screws were used in the ulna for each plate location. After contouring was performed, the plates were removed, and a transverse segment of ulna measuring 3 mm was resected from the nonarticular mid437
438
King etal.
D
J. Shoulder Elbow Surg. November/December 1996
Q
\
Figure 1 Experimentalsetup. Humerus is fixed to Dekin testing jig, ancl cables are secured to brachialis and triceps with heavy suture. Electromagnetic tracking sensors are secured proximal and distal to osteotomy to measure differential motion be~een fragments. portion of the greater sigmoid notch to simulate a comminuted olecranon fracture. By excising a segment of olecranon, no contact occurred between the two fragments, making the model a test of rigidity of the plate/screw construct. The fracture site was not compressed by the fixation, because this would have distorted the greater sigmoid notch, similar to what occurs clinically with comminuted olecranon fractures. The experimental design block randomized the location in which the plate was applied initially for each specimen (Figure 2). Static testing assessed the fixation rigidity of the olecranon by applying equal forces concurrently to the triceps and brachialis tendons from an initial preload of 12.5 N to a maximum of 100 N. This loading range was selected to simulate that which may occur clinically in the postoperative period with early unresisted passive motion. This static stiffness test was performed at 30 ~, 60 ~, 90 ~, and 120 ~ of elbow flexion. After static testing was performed, cyclic loading for 4000 cycles at a frequency of 1 Hz was applied with the elbow at 90 ~ of flexion. This consisted of sinusoidal loading from 25 to 100 N over a period of 67 minutes. The plate location and screws were then changed, and the static and cyclic testing protocol was repeated with the plate and screws in the second position. The ultimate strength of the olecranon fixation was then tested with the elbow at 90 ~ of flexion and incremental loading of the triceps tendon until failure of the fixation or rupture of the triceps
Figure 2 Plating methods. A, Radiograph of lateral plate across decranon osteotomy. B, Radiograph of posterior plate across olecranon osteotomy. tendon suture occurred. The failure test was performed for only one plate location in each specimen, with loading confined to the triceps tendon. A Flock of Birds (Ascension Technology Corporation, Colchester, Vt.) electromagnetic tracking device was used to determine the relative motion of the proximal ulnar fragment to the distal ulna. This device consists of a source that emits an electromagnetic field and sensors that monitor this magnetic field and determine their position and orientation relative to the source. This system provides six degree of freedom kinematic data at 100 Hz. We have determined the mean error of this device to be less than 2% of the measured rotations and translations with a resolution of 0.1 o and 0.25 mm, respectively. ~1 All materials in the field of measurement have previously been demonstrated to have no effect on the accuracy of the tracking system. 11 One sensor was placed on the
J. Shoulder Elbow Surg. Volume 5, Number 6
King etal.
439
Table I Static stiffness Translation (mm)
Rotation (degrees) Angle (~ 30 60 90 120
Plate Poser• ~ral Poser• Lateral Posterior Lateral Pos~rior Lateral
Flexionextension 0.4• 0.8• 0.5• 1.2• 0.7• 0.8• 0.5• 0.8•
Valgusvarus
Internalexternal
Anteriorposterior
Mediallateral
0.2• 0.6• 0.5• 0.4• 0.8• 0.4• 0.6• 0.5•
0.4• 0.8• 0.6• 1.3• 1.2• 1.1• 1.1• 1.4•
0.6• 1.1• 0.5• 1.1• 0.6• 1.0• 0.7• 1.4•
0.3• O.7• 0.4• 0.4• 1.0• 0.4• 0.6• 0.7•
olecranon and one on the distal fragment of the ulna while the transmitter (source) was securely fixed to the testing table. The differential motion between the two fragments was calculated from the data generated by the electromagnetic tracking device. This allowed the determination of implant stiffness in six degrees of freedom. A stylus attached to a third sensor was used to digitize appropriate anatomic markers on the ulna from which a coordinate system was defined. This consisted of digitizing the center of the proximal fracture surface (Point O), the anterior apex of the proximal fracture surface, and the distal center of the ulnar shaft to establish the proximal-distal reference axis. The cross-product of these two vectors (i.e., with a common origin at the center) predetermined the medial-lateral reference axis. Subsequently, the cross-product of this axis with the proximal-distal reference axis established the anteriot-posterior reference axis. The original data from the tracking system was reduced to measure varus-valgus, flexion-extension, and internal-external rotations. The reference frames established for the olecranon and distal ulna were initially superimposed at point O before loading was performed. With the Euler sequence of rotations about the anterior-posterior, medial-lateral and proximaldistal axes of the olecranon, the relative rotation of the reference frame on the olecranon to the distal ulna was determined from basic kinematic algebra of the rotation matrixes. Translations relative to the fracture site were determined by calculating displacement of the olecranon coordinate frame origin from the distal ulnar frame, resolved along the anterior-posterior, medial-lateral, and proximal-distal axes. The static stiffness of the fixation was calculated
Proximaldistal 0.6• 0.7• 0.4• 0.7• 0.7• 0.6• 0.6• 0.6•
as the difference in the relative position of the olecranon to the distal ulna between 12.5 N and 100 N of applied loading. Behavior under cyclic loading was calculated as the difference in relative motion between the initial and final cycle of load application. The block randomized design allowed a comparison of the stiffness and fatigue behaviour of the two methods of plate fixation within the same specimens. The static stiffness data were analyzed with a two-way repeated measures analysis of variance with fixation method and flexion angles as the factors. Plate behavior under cyclic loading was analyzed with a repeated measures one-way analysis of variance. Failure strength was compared with attest. All statistical comparisons were conducted independently for the three rotations and the three translational motions measured, with 0~set at 0.05. RESULTS
Static stiffness testing yielded similar stability at all angles of elbow flexion (p > 0.05) (Table I). No significant difference was seen in the rotations or translations of the proximal olecranon fragment relative to the distal ulna between the two plating methods (p > 0.05). All of the motions were small and near the resolution of the tracking device. Cyclic loading of the specimens resulted in minimal changes in stiffness of the plate constructs over time (Figure 3). No statistical difference was seen in stability between the posterior or lateral plating subjected to cyclic loading (p > 0.05). Rupture of the triceps tendon suture attachment routinely occurred during failure testing at 300 to 500 N of applied force. No gross failure of the hardware occurred in any of the specimens during any phase of the experimental testing protocol.
440
J. Shoulder Elbow Surg. November/December 1996
King et al.
0.5
A
~"
0.4
~
0.3
~ '
0.2
~
0.1
~
0.0
!! F-E
V-V
0.4
--
A
E E
"-" Z iii LLI 0
"
D. (/3 El I-E
030.2-
0.1 0.0---
B
[--7
M-L
A-P
P-D
POSTERIOR LATERAL
Figure 3 Cyclic loading. Magnitude of cyclic stiffness of constructs with elbow held at 90 ~ of flexion for both rotation (A) and translation (B). There were no statistical differences be~een plating methods after 4000 cycles of loading (p > 0.05). FE, Flexion-extension; VV, valgus-varus; IE, internal-external rotation; ML, medial-lateral; AP, anterior-posterior; PD, proximal-distal. Our results demonstrate no significant difference in stability between posterior and lateral plating up to 300 N of triceps tendon loading (p > 0.05). DISCUSSION In this in vitro study we attempted to determine whether posterior and lateral plating of olecranon fractures afforded adequate stability to initiate early protected range of motion. The results of this investigation demonstrate that no significant differences exist in the low load stability of posterior and lateral plating. All of the measured displacements were small and probably not clinically significant. The measured rotations and translations were near the resolution of the tracking device (angulation 0.1 o and displacement 0.25 mm), 11 and therefore smaller differences in stability would not have been detected with this testing protocol. We chose a simulated comminuted fracture model as a "worse case scenario" of fixation stability and measured unconstrained translational and rotational motion at the osteotomy site in six degrees of freedom. The use of elderly cadaveric specimens with poor bone quality suggests that the clinical application of these techniques in younger patients should have even greater strength than reported in this study. Pelvic reconstruction plates were chosen because they are easy to contour to the shape of the olecranon. Although 3.5 dynamic
compression plates have greater rigidity than the pelvic reconstruction plates, they are difficult to contour and are more prominent. Dynamic compression plates may be useful when the olecranon fracture extends into the ulnar diaphysis; however, they are probably not necessary for isolated metaphyseal fractures where the fixation of the screws to bone is probably the weak link. A power analysis was performed to determine the probability of finding a difference in the stability of the two plating methods, if such a difference was present. The minimal clinically significant difference in stability to detect was set at 1 mm and 1~ For a sample size of eight specimens 13 was greater than 0.80 for all measured parameters in both the cyclic and static loading tests. This result suggests that the sample size of the study was sufficient to detect important clinical differences in stiffness between the two plating methods. We selected a level of loading of the brachialis and triceps that would be in the range anticipated for early unresisted active motion during rehabilitation. The tension generated in these tendons with both restricted and normal activity is unknown. Based on published data of muscle moment arms, loading of the triceps tendon to 100 N at a moment arm of 20 mm would balance an anteriorly directed force of approximately 7 N at the wrist (at a length of 300 mm)." 2. 9
King etal.
J. Shoulder Elbow Surg. Volume 5, Number 6
An interesting finding in this study was the somewhat complex and inconsistent motion patterns of the olecranon that emerged for "unidirectional" loading of the triceps tendon. The advantage of an electromagnetic tracking system over that of conventional testing of fixation devices is that it allows unconstrained testing with physiologic loading and assessment of fragment motion in six degrees of freedom. Out-of-plane motions may be an important cause of fixation failure that may go undetected with the more constraining conventional three- or four-point bending tests on a materials testing machine. The ultimate failure loads of the two fixation methods could not be measured because of the failure of the triceps tendon-suture attachment. Whether the absolute failure strength is important in plate fixation is debatable, because clinical failures are usually not catastrophic but rather tend to be due to repetitive loading across the fracture site. We chose cyclic loading as our major factor of interest in an attempt to mimic loading that may occur clinically with early postoperative elbow motion. In certain fracture patterns posterior plating of the olecranon will still be indicated. Patients with displaced coronoid fractures requiring internal fixation should probably have the plate applied posteriorly to allow the coronoid to be lagged back to the plate at the time of the internal fixation. Posterior comminution or extremely small olecranon fragments may also be an indication for posterior plating. Most fractures of the olecranon, however, do not have an associated coronoid fracture, and therefore lateral plating may be used. This method may be particularly indicated for simple olecranon fractures, where a reliable method of fixation is desired with improved softtissue coverage of the hardware. This method may result in a lower incidence of wound complications and postoperative hardware pain than has been reported with tension band wiring or posterior plating. Because of the lateral location of the plate, bicortical fixation of the proximal olecranon fragment may be possible with lateral plating, poten-
441
tially increasing the fixation strength. In this in vitro study both methods of plate fixation afforded adequate stability to permit early protected motion. Before this method can be applied clinically, however, a prospective clinical trial comparing the two plate fixation methods will be required. The authors appreciate the support of Synthes Canada in providing the internal fixation devices used in this study.
REFERENCES
1. An KN, Hui FC, Morrey BF, LinscheidRL, Chao EY. Muscles across the elbow joint: a biomechanical analysis. J Biomechanics 1981 ; 14:659-69. 2. Funk DA, An KN, Morrey BF, DaubeJR. Ebclromyographic analysis of muscles across the elbow joint, J Orthop Res 1987;5:529-38. 3. Fyfe IS, Mossad MM, Holdsworth BJ. Methods of fixation of olecranon fractures: an experimental mechanical study. J BoneJoint Surg Br 1985;67B:367-72. 4. Gartsman GM, Sculco TP, Otis JC. Operative lreatment of decranon fractures. J BoneJoint Surg Am 1981 ;63A;71821. 5. Home JG, TanzerTL. Olecranon fractures: a review of 100 Cases. J Trauma 1981 ;21:469-72. 6. Homer SR, Sadasivan KK, Lipka JM, Saha S. Analysis of mechanical tactors affecting fixation of olecranon fractures. Orthopaedics 1989;12:1469-72. 7. Hume MC, Wiss DA. Olecranon fractures: a clinical and radiographic comparison of tension band wiring and plate fixation. Clin Orthop 1992;285:229-35. 8. Johnson RP, Roetker A, Schwab JF. Olecranon fractures treated with AO screw and lension bands. Orthopaedics 1986;9:66-8. 9. King GJW, Itoi E, Niebur GL, Morrey BF, An KN. Motion and laxily of the capitdlocondylar total elbow arthroplasly. J BoneJoint Surg Am 1994;76A:1000-8. 10. Macko D, Szabo RM. Complications of tension-bandwiring of olecranon fractures. J Bone Joint Surg Am 1985;67A: 1396401. 11. Milne AD, JohnsonJA, ChessDG, King GJW. Accuracyof an electromagnetic tracking device: a study of the optimal operating range and metal interference. J Biomechanics ] 996;29:791-3. 12. Murphy DF, Greene WB, GilbertJA, DameronTB. Displaced olecranon fractures in adults: biomechanical analysis of fixation methods. Clin Orthop 1987;224:210-4. 13. Murphy DF, Greene WB, GilbertJA, DameronTB. Displaced olecranon fracturesin adults: clinical evaluation; Clin Orthop 1987;224:215-23.