Session VII - Foot and Ankle
A Biomechanical Evaluation of Syndesmostic Screw Fixation of Weber C Ankle Fractures
Markku T. Nousiainen, MD1; Alison J. McConnell, BSc2; Mohit Bhandari, MD3; Emil H. Schemitsch, MD2;
Background: Biomechanical and clinical studies have shown that syndesmotic screws are indicated in repair of Weber C ankle fractures if either the fibular fracture is more than 4.5 cm above the tibiotalar joint with an associated deltoid ligament tear or if rigid fixation of bimalleolar fractures cannot be obtained. Screw fixation of the syndesmosis is a rigid repair compared with the natural flexibility permitted by the natural syndesmosis. Because of the geometry of the talus, that is, that the anterior transverse width is greater than the posterior width, there is concern that syndesmotic screws inserted with the ankle in plantarflexion may restrict range of motion. Current screw insertion technique involves keeping the foot in maximal dorsiflexion. One report concluded that insertion in plantarflexion does not restrict range of motion; however, this was tested in an unloaded radiographic study. In addition, there have been no studies that have directly compared the effect the number of cortices of fixation have on ankle joint biomechanics. Therefore, we examined the biomechanical parameters of syndesmotic screw fixation with varied insertion techniques and ankle positions.
Methods: Nine pairs of fresh-frozen cadaveric through-knee amputations were stripped of soft tissue, preserving the knee and ankle joint capsule, interosseus membrane, and ligaments about the ankle. A three-degrees-of-freedom fixture was used to secure the tibia to the crosshead of an Instron 8874 Mechanical Testing Machine (Instron Corp, Canton, Massachusetts). The foot was secured with Steinmann pins through the calcaneus and metatarsals onto a sliding plate. Maximum dorsiflexion was measured with an axial load of 300 N. Diastasis and rotation measurements were made under conditions of no load, axial load of 700 N, and axial load of 700 N plus torsional loads of 1 Nm and 5 Nm. Measurements were made with a 3-D digitizer (Microscribe G2, Immersion Corp., San Jose, California).
Baseline values for maximum dorsiflexion, tibiofibular diastasis, and tibiotalar rotation were obtained. The deltoid ligament and the syndesmosis and interosseus membrane up to 100 mm above the ankle joint were sectioned to simulate a Weber-C pronation external rotation injury. The fibula was left intact to simulate perfect fibular fixation in every specimen, removing this as a source of confounding variance in the data. The right ankles were fixed with the loaded foot in maximum plantar flexion; the left ankles were fixed in dorsiflexion. A 3.5-mm fully threaded cancellous screw was inserted through three cortices of fixation, and maximum dorsiflexion, tibiofibular diastasis, and tibiotalar rotation measurements were repeated. Testing was repeated with another screw inserted to achieve fixation through four cortices. Ankles were axially loaded and externally rotated to failure, and the failure torque and failure angle were recorded. Post-repair values were expressed as changes from baseline values. One-way analysis of variance results were confirmed with multivariable regression analysis.
Results: Diastasis widening was greater with screw insertion in dorsiflexion (P = 0.001 to P = 0.003 for various loading conditions), whereas tibiotalar rotation was not statistically significantly different. Insertion position was able to explain 23% to 31% of the variance in the diastasis measurement under loaded conditions. The diastasis widened when compared with baseline in the dorsiflexion series, with the magnitude of this widening increasing from an average of 0.072 mm ± 1.203, with a pure axial load, to 1.158 mm ± 1.012 with an axial load plus 5 Nm torque. In the plantarflexion specimens, the average diastasis decreased from baseline (-1.179 ± 0.948 mm, under a pure axial load, to -0.456 ± 1.365 mm, with an axial load plus torsional load of 5 Nm). In addition, this change in diastasis did not affect range of motion, as quantified by the change from baseline in maximum dorsiflexion angle (dorsiflexion average, -1.01 ± 2.93 deg; plantarflexion average, -0.51 ± 2.97 deg). The number of cortices of fixation did not significantly affect diastasis, tibiotalar rotation, nor the maximum dorsiflexion angle achieved. Failure stiffness was also not affected by insertion position.
Conclusion/Significance: We report the following: 1) position of the foot upon insertion of syndesmotic screws is independently associated with diastasis widening; 2) this widening does not, however, limit dorsiflexion. This may indicate that, if the ankle joint is anatomically reduced, regardless of the flexion position, the joint will not restrict range of motion.