Minimally invasive reconstruction of lateral tibial plateau fractures using the jail technique: a biomechanical study

Forty fresh skeletally mature porcine tibias were obtained from a local butcher and frozen at −20°C. The porcine model was selected because of the similarities of structural properties between the human and the porcine knees [10]. The specimens were thawed for 12 hours at room temperature and all muscles and soft tissue were removed before testing. After cleaning and degreasing the tibial diaphysis, the lower part was embedded in a metal container using a two-component polyurethane foam with the tibial plateau in an upright position. A standardized osteotomy was performed to simulate fracture of the lateral tibial plateau. The holes for screw osteosynthesis as well as the additional jail screw were drilled before osteotomy. This was to ensure that an exact reposition of the joint surface and the joint congruence after osteotomy was possible. The drill holes had a diameter of 3 mm and were localised 7 mm under the joint line with a 20-mm distance. The osteotomy was performed with an oscillating saw at a distance of 13 mm from the lateral tibial plateau (Figure 2).

Following osteotomy, two different screw fixation techniques for lateral tibial plateau fractures were biomechanically evaluated: the conventional two-screw osteosynthesis (Figure 1a) as described by Büri et al. in 1979 [9] and the jail technique which introduced another screw transverse to osteosynthesis screws as shown in Figure 1b. In the latter, the third screw was placed in an angle of 90° below the two conventional screws in the intact part of the bone next to the fracture site. To ensure that the supporting lower screw is in direct contact with the 2 upper lag screws, a 1-mm K-wire was placed under the two screws and used as a screw guide wire.

For each single specimen, an individual spacer was created using a defined amount of 10 ml bone cement. This spacer was placed right between the punch of the material testing machine and the lateral tibial joint to make sure that an optimal contact surface was achieved. The spacer was made of Palacos® (Heraeus Medical GmbH, Wehrheim, Germany) a two-component bone cement (Figure 2). The whole construct (orientation of the punch, tibial slope, distal tibia) was adjusted in an uniaxial direction to ensure that the conditions at the repair site were repeatable in every individual test setup. This model was chosen after preliminary testing of a control group adjusting the spacer-joint surface interface in an original lateral tibial plateau.

Each individual spacer underwent an axial loading up to 3000 N for 10 minutes prior to testing to confirm that there is no relevant deformation of the spacer in the test setup.

The two tibial fracture reconstructions were subjected to two different loading protocols. In the first protocol, both reconstructions (n = 10 per group) underwent a load-to-failure testing protocol using a material testing machine (Zwick/Roell® Z005, Zwick Gmbh & Co. KG, Ulm, Germany). During load-to-failure testing, an axial load was applied with a loading rate of 200 mm/min on the reconstructed fragment of the lateral tibia simulating a worst case scenario (Figure 2). As a second test, a cyclic loading ramp protocol (n = 10 per reconstruction group) was introduced using an INSTRON® 8874 servo-hydraulic material testing machine (Instron® Deutschland GmbH, Pfungstadt, Germany). A preload of 30 N was first applied to all specimens. All reconstructions were cyclically loaded in 5 steps: 30-200 N, 30-400 N, 30-600 N, 30-800 N and 30-1000 N, 1000 cycles in each step. Cyclic loading was performed at a displacement rate of 200 mm min^-1 and a loading frequency of 80 cycles min^-1. This loading protocol is within the general range reported in previously published studies involving cyclic loading, and represents a relatively modest load level imitating an aggressive rehabilitation protocol [11]. The loading frequency was similar to that of other studies and appears to be within a physiological range of loading [10].

Reconstructions that survived the cyclic loading protocol were finally tested until failure in the same test setup. A total of 40 porcine tibias were used, with each tibia used only for one test setup. All tests were performed at room temperature, and the tibias were kept moist during preparation, mounting, and testing to prevent desiccation with saline.

During the testing process all data was recorded continuously using a computer data recording system. After testing maximum load, yield load, stiffness and displacement were analyzed. Additionally a load displacement curve was recorded. Maximum load was defined to be the highest measurable value of the load displacement curve. Failure (yield load) was defined as the point in the curve where the reconstruction failed and started to undergo plastic deformation.

Stiffness was defined as the slope of the linear region of the load displacement curve.

Displacement was analyzed after the last cycle of each loading step in the ramp protocol (Figure 3) and after loading to failure at the end of every test setup.

Failure mode was monitored by visual analysis, photo and video documentation.

In the single cycle protocol, all reconstructions in both groups tested were loaded until failure. Reconstructions using the conventional osteosynthesis showed a significantly lower maximal load (1769.46 ± 107.47 N) in comparison to the jail technique (2275.87 ± 253.33 N) (p < 0.001).

There was a tendency of better outcomes using the jail technique for stiffness (conventional osteosynthesis = 441.61 ± 98.211 N mm^-1; jail technique = 491.81 ± 98.54 N mm^-1) and yield load values (conventional osteosynthesis 1068.73 ± 110.29 N; jail technique 1154.7 ± 199.92 N) but these results were not statistically significant (p = 0.228, 0.150, respectively).

On the other hand, displacement measurements tended to be lower using conventional osteosynthesis compared to the jail technique but this result did not reach statistical significance (Table 1).

In the second testing protocol, all reconstructions loaded cyclically survived the ramp protocol and were afterwards loaded to failure in the same test setup.

During the cyclic testing, the displacement of the lateral tibial fragment was analyzed. In all 5 steps, reconstructions using the jail technique showed lower displacements in comparison to the conventional osteosynthesis (Table 2). The mean displacement after 5000 cycles was 9.54 ± 1.84 mm for the conventional osteosynthesis and 8.53 ± 1.68 mm for the jail technique. The results, however, did not reach statistical significance.

Looking at the loading to failure after cyclic loading to evaluate the residual stability of the reconstructions, the result showed again a trend for better outcomes in terms of maximal load, yield load, stiffness and displacement for jail technique (Table 3). The differences, however, were not statistically significant.

The failure modes that occurred during the single cycle testing were similar to those observed during the cyclic load testing protocol. Although all reconstructions were loaded to failure, the typical failure modes differed between the two reconstruction groups as shown in Figure 4.

The predominant failure mode in the conventional osteosynthesis group was screws cutting through the bone (cut-outs) of the lateral tibial plateau (92%; Figure 4a). In the course of the experiment, the two lag screws were pushed downwards into the cancellous bone and bent slightly in the direction of force action. At the end of the test, a clear cutting through the tibial spongy bone was observed, an effect that was more evident in specimens tested cyclically than those tested in a single cycle.

In specimens reconstructed with the jail technique, a deformation of the two lateral screws at the site directly above the jail screw was documented as the typical failure mode (85%; Figure 4b) but no cut-outs in the cancellous bone were observed.