2.1 Biomechanical testing

Biomechanical setups were chosen according to previously published studies by Brinkmann et al. who examined lateral open and medial closed wedge techniques of the femur [15–17].

The proximal and distal parts of the sawbones^® were embedded in casting resin (Rencast FC 53, Huntsman Advanced Materials, Basel, Switzerland), with additional socked screws providing rotational stability within the potting on the distal part.

Single plane open wedge osteotomies with medial and lateral plate osteosynthesis using an angle stable tomofix plate were performed according to standard surgical procedures [18]. The following implants were used:

• 4 hole Tomofix lateral distal femur plate (141 mm, Tomofix, DePuy Synthes, Oberdorf, Switzerland) for lateral open wedge procedure.
• 4 hole Tomofix medial distal femur plate (112 mm, Tomofix, DePuy Synthes, Oberdorf, Switzerland) for medial open wedge procedure.

After precise positioning of the osteotomy gap all plates were fixed with eight angle stable locking screws to the composite bone to ensure a reproducible position. The gap size was chosen to be 10 mm in each case. This was ensured by implanting a triangular spacer which was removed after the plate was fixed to the femur. A bony cortex of 5mm length from an anterior perspective was left intact. In order to avoid a fracture of the remaining bone bridge the sawbone^® was slightly heated up during the process of implanting the spacer. This was achieved using a hot air fan. Fig 1 shows the embedded composite bone construct.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Embedded sawbone construct with tomofix plates.

a) medial open wedge procedure; b) lateral open wedge procedure.

https://doi.org/10.1371/journal.pone.0310869.g001

2.2 Testing protocol

Specimens were placed into a loading fixation of a servohydraulic testing machine (Mini Bionix 858; MTS Systems Co, Eden Prairie, USA). The proximal fixation was a universal joint in all cases (compare Fig 1). For axial and ultimate loading, the proximal part was placed in a ball and socket joint. For torsional loading, the proximal part was placed in a universal joint. For all tests, an initial preload of 10 N was applied.

All femurs were subject to an established loading protocol [19]. We applied 100 cycles of 50 to 150 N followed by 100 cycles of 50 to 800 N axial load at 1 Hz in a sinusoidal waveform. Torsional tests were performed with 100 cycles of 5 Nm internal rotational moment to 5 Nm external rotational moment at 0.25 Hz in a sinusoidal waveform under additional axial loads of 0 N, 150 N and 800 N, respectively.The described setup and values were chosen because of their resemblance of a postoperative partial weight bearing protocol following correction osteotomy starting with 15 kilograms and increasing to full weight bearing after four weeks post intervention.

Finally, all specimens were loaded to failure applying axial load at 0.1 mm/sec up to 8000 N. Premature failure was assessed macroscopically and defined as either fracture of the bony cortex or implant failure. A microscopical examination for possible damage to the osteotomy or implant was not performed.

Load and displacement of the actuator was continuously monitored by the testing machine at 1000 Hz.

2.3 Analysis

Axial stiffness was calculated for each loading cycle by dividing maximum force by maximum relative displacement. Torsional stiffness was calculated by dividing torque range by displacement range for each cycle. For the final analysis, all cycles of a loading step were averaged.

The Mann—Whitney-U-test for independent samples was used to compare stiffness of each loading mode and load-to-failure tests between plate setups. An alpha level of 0.05 was used as threshold for significance.

The sample size was determined by orientation on comparable biomechanical studies using composite bones, which exhibit lower variability of mechanical properties compared to human tissue.

3.1 Main findings

No damages to the Sawbone^® constructs were evident throughout axial and torsional testing. No statistically significant difference could be noted between testing groups regarding most parameters.

Both experimental groups proved stable under axial and torsional loadings. Significant differences were noted for torsional stiffness under low and mid force loadings (P = 0.002; P = 0.009), favoring the medial open wedge constructs (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Overview of results for axial and torsional loadings for both testing groups.

https://doi.org/10.1371/journal.pone.0310869.t001

3.2 Axial stiffness

No evident damage to the bone or implant was noted during cyclic loadings (Table 1). During axial loading protocol no statistically significant differences could be noted under low- and high force loadings (Fig 2). It should be noted that there is a tendency for an increased axial stiffness on the medial open wedge procedure.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Axial stiffness under high force loading (800N).

https://doi.org/10.1371/journal.pone.0310869.g002

3.3 Torsional stiffness

No evident damage to the bone or implant was noted during cyclic loadings (Table 1). Torsional stiffness was significantly greater in the medial open wedge group under low (0 N P = 0.002) and mid force (150 N; P = 0.009) loadings. No significant difference could be noted under high force loadings.

3.4 Failure mode

Failure mode for all failed specimens (lateral: n = 1, medial n = 2) was the breakage of the bony cortex. All other implants yielded the maximum load of 8000 N.

4.1 Limitations

Several limitations apply to this study. First, being a biomechanical study, the biological healing process and postoperative rehabilitation, which are crucial to the surgery’s outcome, could not be taken into account.

Second, we can only provide the time-zero data of biomechanical stability following osteotomy and locking plate osteosynthesis. In addition, due to the use of composite bones there is no biological or anatomical variability to be found among the constructs.

Moreover, in this examination composite bones instead of human femurs were used. Even though their stiffness values are supposed to resemble the ones of natural human bones, our experiment required slight preheating of the construct in order to be able to create an osteotomy gap of ten millimeters without producing an immediate hinge fracture. Therefore, we see a need for further in-vitro examinations, ideally using full-length human femurs to confirm the data acquired from this composite bone study. This is emphasized by the fact that to our knowledge there are no other studies examining a medial open wedge procedure on the distal femur.