The study was approved by the Medical Research Ethics Committee of Acibadem Mehmet Ali Aydinlar University (Approval No: 2019-10/4). A total of 20 wrists from 10 fresh frozen cadavers were supplied from Acibadem University. The cadavers were eligible if they had had no tumor, bone defect or lesion, osteoarthritis, previous fracture, or osteoporosis. We prepared homogeneous groups according to the dominant and non-dominant wrists of the cadavers.
We separated radius bones from the ulna and other soft tissues and cut into specimens of 14-cm in length. Afterwards, we placed TST® titanium alloy distal radius anatomical plates to the distal radii in full anatomical position, just proximal to the watershed line. Considering the count and configuration of the screws suggested by Mehling et al. , we sent three bicortical screws to the shaft of the radius, followed by unicortical drilling for distal screwing. We performed measurements by pulling the drill once it reached the opposite cortex.
We selected the screw lengths such that they corresponded to the 75% of the measured length (short-length). In the same configuration for each of the cadavers, we delivered six screws from distal radius holes of the anatomical plate. Afterwards, we used oscillating handsaw to create extraarticular distal radius fracture model (AO 23-A3.2) [4, 14]. We created dorsal ap model by performing a 1-cm wedge osteotomy from dorsal aspect (Fig. 1, 2). Complete separation of the volar cortex was achieved. Potting was performed by embedding the shaft of the prepared radius into the polyurethane medium. We placed aluminum apparatus into the distal end to ensure applying of torsional and axial loading in biomechanistic tests (Fig. 3, 4).
All specimens were placed in the testing machine and tested under axial and torsional loads. Initially, axial and torsional forces were simultaneously applied to each sample, where stiffness and elastic limit measurements were obtained. Afterwards, the magnitude of maximal force that was required to achieve catastrophic failure (fracture of the bone, screw, or plate) were determined for both short-length and full-length groups.
The plated cadaveric radius bones were embedded into the polyvinyl chloride tube from one end via polyester resin, while the other end was fixed to the test device via a miniature vise. This vise ensured both torsional and axial compression by clamping the bone in a plane perpendicular to the plate plane.
A vise was attached to the loading cell (AXIAL-TORSIONAL LOAD TRANSDUCER 25 kN / 25 Nm) of the testing device (MTS 858 Mini Bionix II), and a steel pot was placed in this vise to place the samples. The prepared samples were placed inside the steel pot via PVC tube. The upper part of the bone was attached to the test device through the miniature vise. While the loads applied to the bone were measured via the transducer (AXIAL-TORSIONAL LOAD TRANSDUCER (2500 N / 25 Nm), the displacements and angles were calculated via the displacement transducer (MTS LVDT TRANSDUCER-359/LVDT, Displacement, Serial Number: 10188729) and angle transducer (MTS ADT TRANSDUCER- 359/ADT, Torsional Angle, Serial Number: C11382).
Once the test began after connection of the samples to the device, a torsional load of 0.5 Nm to 5 Nm was applied for 10 cycles simultaneously with an axial compression force between 5 N and 250 N at a frequency of 0.25 Hz to eliminate the gaps in the system and observe the range in which the system operates stably. Afterwards, the loads in the system were reset and the axial and torsional stiffnesses of the system and the maximum loads that the system can carry were determined with static loading. These static tests were performed with an axial speed of 2 mm/min and a rotational speed of 10°/min. The occurrence of closure of the osteotomy line or the loosening of the screws were accepted as damage criteria, upon which the tests were terminated.
Axial and torsional stiffnesses of the samples under static loadings, their elastic limits, as well as the axial compression and torsional moments detected at the moment of fracture were calculated using MATLAB 2018 software.
We used Number Cruncher Statistical System 2007 (Kaysville, Utah, USA) software for statistical analysis. Continuous parameters were expressed as mean, standard deviation, median, minimum, and maximum. For the non-normally distributed data, we compared the groups through Mann-Whitney U test. An overall Type-I error level of five percent was used to infer statistical significance.