ORIGINAL RESEARCH

Syndesmosis Fixation: Analysis of ShearStress via Axial Load on 3.5-mm and 4.5-mm Quadricortical Syndesmotic Screws

Matthew Hansen, DPM,1 Long Le, DPM,2 Stuart Wertheimer, DPM,3 Eric Meyer, BS,4 andRoger Haut, PhD5

The effect of shear stress on a fixated distal syndesmosis of the ankle was evaluated with a servohydraulicmaterials–testing machine. Eighteen syndesmoses were fixated in a quadricortical technique using3.5-mm cortical and 4.5-mm cortical stainless steel screws. A shear stress was applied via an axial loadin an attempt to simulate weightbearing. The 4.5-mm quadricortical screws produced a higher yield loadand peak load (484.3 � 93.8 N and 597.7 � 81.4 N) when compared with the 3.5-mm quadricorticalsyndesmotic screws (412.8 � 55 N, P � .033 and 511.2 � 64.4 N ). These findings suggest that a largerdiameter screw provides greater resistance to an applied shear stress at the distal syndesmosis. (TheJournal of Foot & Ankle Surgery 45(2):65–69, 2006)

Key words: syndesmosis, syndesmotic fixation, distal tibiofibular ligaments

Syndesmotic injuries of the ankle commonly occur via anexternal rotational force applied to the ankle joint. Injuriesto the distal tibiofibular ligaments occur in pronation-exter-nal rotation fractures, supination-external rotation fractures,and Maisonneuve fractures (1, 2). Disruption of the syndes-motic ligaments can also occur with severe ankle sprainswithout fractures, which are termed high ankle sprains(3–5).

There are many surgical approaches for placement ofsyndesmotic fixation in reference to type and number of

Address correspondence to: Matthew P. Hansen, DPM, St. John NorthShores Hospital, 26755 Ballard Road, Harrison Township, MI, 48045.E-mail mphansen43@hotmail.com

1Submitted while third year resident, Podiatric Surgical Residency, St.John North Shores Hospital, Harrison Township, MI.

2Submitted while second year resident, Podiatric Surgical Residency,St. John North Shores Hospital, Harrison Township, MI.

3Director of Podiatric Education, St. John North Shores Hospital, Har-rison Township, MI.

4Submitted while graduate student, Mechanical Engineering, College ofEngineering, Michigan State University, East Lansing, MI.

5Director and Professor, Orthopedic Biomechanics Laboratories, Col-lege of Osteopathic Medicine, Michigan State University, East Lansing,MI.

Copyright © 2006 by the American College of Foot and Ankle Surgeons

1067-2516/06/4502-0002$32.00/0doi:10.1053/j.jfas.2005.12.004

V

fixatives, number of cortices, and level of fixative insertion(6–9). Guidelines recommend placement of the syndes-motic screw between 2.0 cm and 3.0 cm superior to thetibial plafond, parallel to the ankle joint, and with 20 to 30°angulation from the frontal plane (10). Fully threaded3.5-mm or 4.5-mm cortical screws are inserted perpendic-ular to the tibiofibular joint with the neutralization tech-nique, not the lag technique (11, 12).

Debate exists whether the screws should purchase 3 or 4of the available cortices of the tibiofibular joint. Proponentsof the tricortical technique suggest that this fixation does notalways require removal before full weightbearing (12, 13).The observed presence of micromotion suggests that thesescrews are more likely to loosen rather than fracture. Re-sorption of the bone surrounding the tricortical screws mayhelp reestablish normal motion within the syndesmosis, thusmaking routine removal of the tricortical screws unneces-sary (14). Others suggest that the quadricortical technique ismore rigid, with less occurrence of syndesmotic wideningduring healing, and easier removal if screw failure occurs(6, 10, 15, 16).

There is also no consensus regarding weightbearing sta-tus after a repair of the syndesmosis with screw fixation.Many believe weightbearing is only allowed after syndes-

motic screw removal, because syndesmotic fixation has

OLUME 45, NUMBER 2, MARCH/APRIL 2006 65

been shown to limit tibiotalar external rotation, increasetalar inversion, and impart compressive strains on lateralaspect of the fibula (17–19). But a study by de Souza et aladvocated leaving syndesmotic screws intact with earlyweightbearing. They demonstrated that patients had accept-able outcomes without syndesmotic screw removal beforeweightbearing (13).

To our knowledge, there are no studies that compare asmaller versus larger diameter syndesmotic screw with thequadricortical technique with regard to syndesmostic stabil-ity. The purpose of this study was to compare 3.5-mm and4.5-mm quadricortical screws when the syndesmosis wassubjected to a simulated weightbearing force of shear stress.The authors hypothesized that there would be a clinicallysignificant difference in stability between these 2 constructs.

Materials and Methods

Eighteen ankle sawbone specimens consisting of a distaltibia and distal fibula (Pacific Research Laboratories, Va-shon, WA) were obtained for this study and separated into2 groups. Group A consisted of 9 specimens fixated withone 3.5-mm quadricortical stainless steel screw across thedistal syndesmosis (Synthes, West Chester, PA).

Group B consisted of 9 specimens fixated with one4.5-mm quadricortical stainless steel screw across the distalsyndesmosis.

All procedures were performed by the same investigator(L. L.) to provide consistent screw placement. To allow fora consistent angle of insertion, a 0.045-in Kirschner wirewas placed 2 cm superior and parallel to the tibial plafond.The wire was placed from the fibula into the tibia, 30°anterior to the frontal plane. This angle was confirmed witha protractor. The Kirschner wire was removed, and theappropriate-sized drill was inserted along the same guidehole. A 2.5-mm drill and 3.5-mm tap were used for the3.5-mm screw, and a 3.2-mm drill and 4.5-mm tap wereused for the 4.5-mm screw. Each screw was inserted in aquadricortical fashion, with purchase of 2 threads beyondthe medial tibial cortex. The screws were tightened to a firmtwo-finger tightness. The screw length was 50 mm, whichwas determined by a depth gauge.

A cylindrical potting base was created (by E. M.) ofpolymethylmethacrylate (Fibre Strand Body Filler 6371,Cleveland, OH) to allow for stabilization of the distal fibuladuring loading. Before hardening of the potting material, thedistal fibula was partially inserted into the potting cylinder.Once the potting material became firm enough to keep itsshape, the fibula was removed and the potting material wasallowed to harden completely. The potting cylinder wasthen attached to the hydraulic actuator.

Each specimen was tested until failure by loading in a

servohydraulic materials testing machine (Model #1331;

66 THE JOURNAL OF FOOT & ANKLE SURGERY

Instron Corp., Canton, MA). Load was transmitted to thelateral malleolus through the custom-fitting polymethyl-methacrylate interface along the axis of the fibula. The tibiawas constrained to the table of the testing machine with a1.27-cm (0.5-in) diameter steel rod that was inserted 12 cminto the tibial diaphysis to maintain axial alignment (Fig 1).A 2-cm section of the midshaft of the fibula was removed toallow proximal displacement of the fibula relative to thetibia without contacting the table. The fibula was uncon-strained and axially loaded to produce shear stress acrossthe stabilized syndesmosis. The combination of the force onthe fibula and constraint of the tibia produced a simulatedweightbearing shear stress across the distal syndesmosis.

Experiments were performed in displacement control,with the actuator moving a total of 20 mm in 6 seconds

FIGURE 1 (A) Photograph of testing apparatus and specimenprepared with a 3.5-mm syndesmotic screw. (B) Diagrammatic rep-resentation of testing apparatus.

during the loading phase. Load was recorded at 200 Hz with

raph

an 11.1-kilonewton load transducer (Model #10101a-2500;Instron Corp.) that was attached between the actuator andthe polymethylmethacrylate interface. Load-displacementdata were recorded by a personal computer equipped with a16-bit analog/digital board (Model DAS 1600; ComputerBoards, Mansfield, MA). After each experiment, the modeof failure for the complex was recorded. The initial yieldload and displacement were found by locating the point onthe load-displacement graph where the slope (stiffness) sig-nificantly decreased. The energy (area under the load-dis-placement curve) and stiffness up to the initial yield loadwere also calculated. After the initial specimen yield, therewas a region of plastic deformation in either the screw orsawbone and a second point of interest, the ultimate load. Inaddition, the ultimate load, displacement at ultimate load,

0

100

200

300

400

500

600

0L

oad

(N

)

Stiffness

FIGURE 2 Representative load-dis-placement graph of a 3.5-mm diam-eter screw fixation (group A).

0

100

200

300

400

500

600

0150

Disp

Lo

ad (

N) Yield Point

Stiffness

P

FIGURE 3 Representative load-displacement g

and the total energy absorbed before gross failure were

V

documented. The mean and standard deviations of theseparameters were computed within each group and comparedstatistically with the Student t test. Statistical significancewas indicated at P � .05.

Results

A representative graph of load versus displacement wasplotted for one specimen from each group in Figures 2 and3. The initial linear slope represents increasing load withoutpermanent deformation in the screw or sawbone specimen.After this initial phase, there is a nonlinear region, duringwhich permanent structural deformation occurred. In groupA (3.5-mm screw), the deformation consisted of permanent

10 15 20 25

Displacement (mm)

eld Point

Ultimate Load

Plastic Deformation Region

520251

ent (mm)

Ultimate Load

Deformation Region

of a 4.5-mm diameter screw fixation (group B).

5

Yi

lacem

lastic

bending of the screws in 7 of 9 specimens. The final

OLUME 45, NUMBER 2, MARCH/APRIL 2006 67

characteristic of this graph was a sharp decrease in the loadas fracture occurred in the fibula of 6 specimens (Fig 4). Ingroup B (4.5-mm screw), there was permanent screw bend-ing in only 2 specimens and fibular fractures in 4 specimens,

FIGURE 4 Test specimen showing failure mode of 3.5-mm syn-desmotic screw. Note the fracture of the fibula and the bending ofthe screw.

TABLE 1 Experimental data and failure descriptions of specime

SpecimenNumber

YieldLoad (N)

YieldDisplacement

(mm)

Stiffness(N/mm)

Energy toYield(Nm)

A1 350 3.6 93.12 165.7A2 475 6 78.86 599.4A3 463 8.3 53.17 1171.7A4 387 8.5 47 644.7A5 466 9 54.5 1609.8A6 378 5.8 65.51 268.4A7 349 5 70.59 194.5A8 473 8.3 52.9 728.2A9 374 6.43 59.08 320.7Average 412.8 6.8 63.9 633.7Std dev 55.0 1.9 14.8 486.8B1 555 7.7 72.25 725.6B2 439 7.1 64.25 464.2B3 459 6 75.87 384.3B4 377 5.9 62.98 304.2B5 492 8.8 56.68 811.0B6 403 5.8 71.09 312.4B7 549 5.3 103.1 350.8B8 413 6.6 62.82 418.5B9 672 7.6 90.04 859.7Average 484.3 6.8 73.2 514.5Std dev 93.8 1.1 14.8 221.4Student t

test 0.033 0.492 0.099 0.257

Abbreviations: std dev, standard deviation; sb, screw bent; tc, tibiaout of fibia.

68 THE JOURNAL OF FOOT & ANKLE SURGERY

thus most of the plastic deformation in this group was dueto compression near the screw holes in the tibia and fibula.

The initial yield load, yield displacement, energy to yield,stiffness, ultimate load, displacement at ultimate load, en-ergy to failure, and the description of failure were obtainedon all 18 specimens (Table 1). Although both groupsshowed linear regions initially, the yield load was signifi-cantly higher for group B (484.3 � 93.8 N) than group A(412.8 � 55 N, P � .033). There were no statistical differ-ences between the yield displacement, energy to yield, orstiffness of the 2 groups; however, stiffness was slightlyincreased for the 4.5-mm screw (73.2 � 14.8 N/mm vs. 63.9� 14.8 N/mm, P � .099). The ultimate load of 597.7 �81.4 N in group B was significantly higher than the ultimateload of 511.2 � 64.4 N in group A (P � .012). However,there was not a difference in the displacement at ultimateload or the total energy absorbed.

Discussion

This study analyzed the effect of shear stress across thesyndesmosis fixated with 3.5-mm and 4.5-mm quadricorti-cal screws. The yield load and ultimate load data show4.5-mm syndesmotic screws resist the shear stress of sim-

r 3.5-mm (group A) and 4.5-mm (group B) screw diameters

mated (N)

Displacementat UltimateLoad (mm)

Total EnergyAbsorbed

(Nm)

Failure Description

.8 12 2824.8 sb, tc, F7.2 960.3 fc, fp

.9 18 6523.6 sb, tc, fc, fp

.8 18.9 4725.5 sb, tc, F, fc

.2 17.1 4045.3 sb, tc, F

.5 15.2 2795.9 sb, tc, F, fc

.3 12.6 1645.0 sb, fc, fp

.4 15.3 3326.3 sb, tc, F, fc

.9 15.3 2830.9 tc, F

.2 14.6 3297.5 7 sb, 7 tc, 6 F, 4 fc, 3 fp

.4 3.6 1654.3

.9 14.3 3293.7 tc, fc

.2 14 2722.4 F, fc

.5 17.3 4300.6 tc, fc

.1 15.7 3455.3 tc, F20 5755.6 tc, fc

.4 20 5765.7 tc

.1 10.7 2105.2 sb, tc, F

.7 14.5 2836.1 tc, F

.2 10 5250.8 sb, F, fc

.7 15.2 3942.8 2 sb, 7 tc, 4 F, 5 fc

.4 3.5 1378.3

.012 0.375 0.191

pressed; F, fibula fracture; fc, fibula compressed; fp, screw pulled

ns fo

UltiLoa

48249555159258050039555244951164

58261351558053753175055871059781

0

com

ulated weightbearing better than 3.5-mm syndesmoticscrews.

During weightbearing, cyclic and shear loads are trans-mitted through the distal syndesmosis (20). Cyclic loadingmay cause destabilization of the fixation site by fatiguefailure of the screw, loosening of the screw, and distortionof the screw hole (21). If weightbearing were allowedpostoperatively, the stability of the repair may influence theultimate position and integrity of the syndesmosis. Finsen etal concluded that there were no adverse effects for patientswho underwent syndesmotic screw fixation and wereweightbearing in a below-the-knee cast and rubber walkerpostoperatively (22). The findings of the current study sug-gest that 4.5-mm quadricortical screw stabilization of thesyndesmosis may resist disruption with load applied to thefibula in the early postoperative period. However, our studydid not assess cyclic loading.

Because of the use bone models, the stabilizing effects ofthe muscular, capsular, ligamentous, and tendinous struc-tures are not taken into account. However, Landsman andChang ascertained the validity of using bone models toassess relative stability (23).

Other limitations of this study involved specimen prepa-ration. A torque screwdriver was not used to insert thesyndesmotic screws. Even though the screws were insertedin a similar fashion by the same investigator, consistencyamong specimens may have been compromised. Screw in-sertion was also performed based on visual inspection in-volving placement of the syndesmotic screw. Visual inspec-tion may have allowed for small differences in syndesmoticscrew placement.

The results of this limited experimental study suggest thata single 4.5-mm quadricortical screw provides statisticallyimproved resistance to shear stress when applied to thedistal syndesmosis. When early postoperative stresses areanticipated, the authors suggest that a 4.5-mm cortical syn-desmotic screw may resist shear stress better than a 3.5-mmcortical syndesmotic screw.

References

1. Lauge-Hansen N. Fractures of the ankle. II. Combined radiographicand experimental-roentgenologic investigations Arch Surg 60:957–985, 1985.

2. Mosier-LaClair S, Pike H, Pomeroy G. Syndesmosis injuries: acute,chronic, new techniques for failed management. Foot Ankle Clin7:551–565, 2002.

3. Teitz C, Harrington R. A biomechanical analysis of the squeeze test forsprains of the syndesmotic ligaments. Foot Ankle Int 19:489–492,

1996.

V

4. Clanton T, Paul P. Syndesmosis injuries in athletes. Foot Ankle Clin7:529–549, 2002.

5. Hopkinson W, St. Pierre P, Ryan J, Wheeler J. Syndesmosis sprains ofthe ankle. Foot Ankle Int 10:325–330, 1990.

6. Xenos J, Hopkinson W, Mulligan M, Olson E, Popovic N. Thetibifibular syndesmosis. J Bone Joint Surg 77A:847–856, 1995.

7. Tornetta P, Spoo J, Reynolds F, Cassandra L. Overtightening of theankle syndesmosis: is it really possible? J Bone Joint Surg 83A:489–492, 2001.

8. Moseheiff R, Liebergall M, Marguiles J, Peyser A, London E, Segal D.Technical complications of the tibiofibular syndesmotic screw. J FootAnkle Surg 32:462–466, 1993.

9. Olerud C. The effect of the syndesmotic screw on the extensioncapacity of the ankle joint. Arch Orthop Trauma Surg 104:299–302,1985.

10. Muller ME, Allgower M, Schneider R, Wilenegger H. In Manual ofInternal Fixation: Techniques Recommended by the AO-ASIF Group,ed 3, pp 610-611, Springer-Verlag, New York, 1991.

11. McBryde A, Chiasson B, Wilhelm A, Donovan F, Ray T, Bacilla P.Syndesmotic screw placement : a biomechanical analysis. Foot AnkleInt 18:262–266, 1997.

12. Heim D, Heim U, Regazzoni P. Malleolar fractures with ankle jointinstability-experience with the positioning screw. Unfallchirurgie 19:307–312, 1993.

13. de Souza L, Gustilo R, Meyer T. Results of operative treatment ofdisplaced external rotation-abduction fractures of the ankle. J BoneJoint Surg 71A:23–27, 1989.

14. Hoiness P, Stromsoe K. Tricortical versus quadricortical syndesmosisfixation in ankle fractures: a prospective, randomized study comparingtwo methods of syndesmosis fixation. J Orthop Trauma 18:331–337,2004.

15. Henley M, Law M, Henry S. Removal of broken syndesmotic screws:a technical tip. Orthopedics 12:643–644, 1994.

16. Wuest T. Injuries to the distal lower extremity syndesmosis. J AmAcad Orthop Surg 5:172–181, 1997.

17. Ebraheim N, Mekhail A, Gargasz S. Ankle fractures involving thefibula proximal to the distal tibiofibular syndesmosis. Foot Ankle Int18:513–521, 1997.

18. Needleman R, Skrade D, Stiehl J. Effect of the syndesmotic screw onankle motion. Foot Ankle 10:17–24, 1989.

19. Sinisaari I, Luthje P, Mikkonen R, Ruptured tibio-fibular syndesmosis:comparison study of metallic to bioabsorbable fixation. Foot Ankle Int23:744–748, 2002.

20. Cox S, Mukherjee D, Ogden A, Mayeux R, Sadasivan K, Albright J,Pietrzak W. Distal tibiofibular syndesmosis fixation: a cadaveric, sim-ulated fracture stabilization study comparing bioabsorbable and me-tallic single screw fixation. J Foot Ankle Surg 44:144–151, 2005.

21. Thompson M, Gesnik D. Biomechanical comparison of syndesmosisfixation with 3.5 and 4.5 millimeter stainless steel screws. Foot AnkleInt 21:736–741, 2000.

22. Finsen V, Saetermo R, Kibsgaard L, Farran K, Engebresten L, BolzKD, Benum P. Early postoperative weight-bearing and muscle acitiv-ity in patients who have a fracture of the ankle. J Bone Joint Surg71A:23–27, 1989.

23. Landsman A, Chang T. Can synthetic bone models approximate themechanical properties of cadaveric first metatarsal bone. J Foot Ankle

Surg 37:122–127, 1998.

OLUME 45, NUMBER 2, MARCH/APRIL 2006 69