Retrograde Tibial Nailing: a minimally invasive and biomechanically superior alternative to angle-stable plate osteosynthesis in distal tibia fractures

For the surgical feasibility study of the RTN, three fresh cadaveric lower extremities with no previous medical history related to the investigated limbs were obtained from the Anatomical Institute of our University. The use of human cadaveric specimen was approved by the local ethics committee (reference number 837.088.07, 28 March 2007). The RTN is an experimental intramedullary nail under design by our research group. The nailing system features double proximal and triple distal locking options. The nail of 8 mm in diameter and 120 mm in length was implanted into three intact human lower limbs including all soft tissues. Three different fracture types (AO/OTA 43 A1-3) were simulated by osteotomy via a posterolateral approach. After reduction, a 2–2.5-cm-long skin incision was started 1 cm proximally to the tip of the medial malleolus and was extended distally. The tip of the malleolus was marked with a Kirschner wire under fluoroscopic control in anteroposterior (a.p.) and lateral view. The wire was drilled parallel to the medial cortex at a distance of approximately 5 mm. The cortex of the malleolus was opened and a path was created by a sharp 9-mm cannulated awl. The awl was advanced with low force and twisting motions until the medullary canal was reached. The awl and wire were then removed. The nail, assembled on the aiming device, was introduced with low force and small twisting movements until its end sat flush with the entry portal. The nail position was confirmed in a.p. and lateral fluoroscopic views. The trocar combination was inserted until it touched the bony surface and the trocar was removed. Drilling was performed with a 3.2-mm drill bit to the desired depth. In all positions except for the second most distal locking position, which directs at the distal tibiofibular joint, bicortical drilling was carried out. Distal locking was achieved with 4.0-mm dual-core screws, specially designed for optimized purchase in cancellous bone. Proximally, the RTN was locked with standard 4.0-mm cortical screws. Finally, an end cap was introduced at the nail end to create a fixed-angle construct with the most distal screw. All implantations were carried out by the same orthopedic surgeon.

We conducted biomechanical testing to investigate the properties of the two implant types. Our null hypothesis was that there is no difference in the biomechanical properties after RTN osteosynthesis compared to angle-stable plate osteosynthesis (Medial Distal Tibial Plate, Synthes^®, West Chester, PA, USA) in an extra-articular distal tibial fracture model (AO/OTA 43 A3)[18]. Fourth-generation biomechanical composite bone tibiae (Sawbones Europe, Malmö, Sweden, item number 3401) were used for the experiments. Based on a priori power analysis, 16 composite tibiae, 8 in each group, were tested in the comparative study. Implantations of the four-hole Medial Distal Tibial Plate (MDTP) were performed as stated in the manufacturer's instructions using three proximal and four distal screws. First, non-angular stable screws were used to fix the plate to the tibia proximally and distally. Afterwards, the two angle-stable screws were drilled and introduced proximally and three distally. The RTN was implanted as described above. All implantations were carried out by the same orthopedic surgeon.

After implantation into the intact composite bone, the osteotomy parameters were established. A transverse defect osteotomy of 10 mm was performed to simulate an AO/OTA 43 A3 type fracture. The bone between 40 and 50 mm from the tibial plafond was resected by a parallel saw. The proximal and distal segments were potted in poly-methyl methacrylate (PMMA) with two screws at each interface to provide a secure force transmission. To ensure that no direct contact between the PMMA and implants existed, all protruding parts were covered with modeling clay prior to potting. Additionally, all samples underwent radiological control in a.p. and lateral view to exclude possible damages and to document the correct implantation.

A universal pneumatic testing machine (SincoTec Test Systems, Clausthal-Zellerfeld, Germany) controlled by PneuSys software (SincoTec, Clausthal-Zellerfeld, Germany) was used for testing. The setup was based on the method described by Hansen et al.[19]. For extra-axial compression, the point of load transmission was in accordance to Horvitz et al.[20]. The force was applied to the tibial plateau at a 10-mm posteromedial offset from the tibial eminentia. A steel ball in a trough coupled with a force-transmitting bar applied the pressure to the plateau. Distally, a cardan joint was used to position the samples in the testing machine since the human ankle joint is best simulated by this method.

The axial loading began under a constant preload of 18 N and was increased to 350 N before decreasing back to 18 N. Each cycle had duration of 20 s. After a pre-cycle, three measurement cycles were recorded. During that phase, the actuator connected to the force-transmitting bar of the testing machine recorded axial movements.

For torsional testing, the specimens were anchored at both ends. Bidirectional torque was applied under a constant preload of 10 N. A cycle consisted of torsion from 0 to 8 Nm clockwise and back to 8 Nm counter-clockwise with duration of 20 s per cycle. One pre-cycle and three measurement cycles were conducted in this study. During that time a rotational transducer connected to the actuator of the testing machine recorded the rotational difference of the proximal to the distal segment.

As a final investigation, an axial ‘load-to-failure’ test was performed for extra-axial compression of up to 1,200 N.

Construct stiffness was measured during axial loading tests by the load to deformation curve. Rotational stiffness was measured by the torsion to rotation curve. The correlation of the constructs' linear-elastic area of the force-displacement and force-angle diagram was the basis for the non-destructive test evaluations. Force-failure values were recorded and analyzed for load-to-failure test. Failure was defined as plastic deformation of the construct. All constructs underwent the identical testing sequence. First, a non-destructive test for extra-axial force with 350 N was performed. Afterwards, a bidirectional rotational test with 8 Nm was done. Finally, increased loading for destructive loading tests took place for extra-axial compression up to 1,200 N. Table 1 gives an overview of the testing sequence.

An a priori power analysis was performed to determine the sample size (eight in each group) for the biomechanical testing. Statistical analysis was performed with SPSS software (SPSS 21.0, IBM, Armonk, NY, USA). The force-displacement and force-angle values of the two implant types were compared with the Student's t test. An error level of less than 5% (p value, 0.05) was defined as statistically significant.

Retrograde intramedullary nailing with the newly developed RTN was possible in a lab setting without major complication through limited skin incisions of 2–2.5 cm for the nail insertion and stab incisions for interlocking. When the entry portal is chosen on the medial surface of tip of the medial malleolus and the trajectory parallel to the medial cortex, the risk of damaging or fracturing the medial malleolus is very low. In our study, this did not happen but it is a complication which has to be taken into account. Moreover, no iatrogenic posterior tibial tendon injury occurred. In simple fracture types, closed manual reduction with the aid of percutaneous reduction forceps is possible. In highly comminuted fractures like the AO/OTA 43 A3, the use of a temporary large distractor may be useful to achieve correct axis and rotation prior to intramedullary nailing. Otherwise, primary malalignment can result. Figure 1 shows an implanted RTN in a cadaveric lower limb with an AO/OTA 43 A3 fracture. Figure 2 demonstrates the limited skin incisions that are necessary for implantation of the RTN.

Results show a higher stability for the RTN during clockwise and counter-clockwise rotations. The average force-angle value was 1.83 ± 0.33 Nm/° in the RTN vs. 0.39 ± 0.03 Nm/° in the MDTP group during clockwise rotation (Figure 4a) respectively, 1.83 ± 0.26 Nm/° in the RTN vs. 0.55 ± 0.08 Nm/° in the MDTP group during counter-clockwise rotation (Figure 4b). Statistical analysis proved a significant difference (p values < 0.001) between the RTN and MDTP group for torsional stability.

Destructive extra-axial compression resulted in failure of all MDTP constructs with an average value of 375 ± 73 N for plastic deformation and the fracture gap closed at an average value of 561 ± 63 N. Due to the extensive deformation, the load increase was stopped at 900 N in the MDTP group. The Kaplan-Meier survival curves for destructive extra-axial compression in the form of plastic deformation and closure of the fracture gap are shown in Figure 5. All RTN specimens survived the maximal load of 1,200 N, which was limited by the maximal force applied by the testing machine. Plastic deformation was not recorded in any RTN specimen. Images of the destructive extra-axial compression test are given in Figure 6.