Comparisons of external fixator combined with limited internal fixation and open reduction and internal fixation for Sanders type 2 calcaneal fractures

Finite element analysis: collection of imaging data

A GE 64-row spiral CT scanner (Siemens, Munich, Germany) supported by the Southern Medical University Fengxian District Central Hospital was used to conduct the helical scan of the male volunteer’s right lower limbs (173 cm height and 64 kg weight, the trial had been approved by the hospital’s ethic committee, and an informed consent had been signed), and the data were saved in Digital Imaging and Communications in Medicine format. The scanning conditions were as follows: 155 mAs at 120 kV. The thickness from 15 cm above the ankle to the planta pedis was 2 mm. The size of the pixel matrix density in each scanning layer was 512 pixels × 512 pixels.

Finite element model of calcaneus construction

The data were inputted into the Mimics 15.01 software (Materialise NV, Leuven, Belgium). A computer-aided design (CAD) model (Figs 1 and 2) was generated by a method proposed by He et al.⁶ The model was wrapped and smoothed through the functions of the 3D reconstruction, including region growing, cavity filling and editing masks, to obtain the desired 3D finite element model of the lower extremity. A Sanders type 2 fracture model was created using the cutting tools of SolidWorks 2012 (Dassault Systemes, Boston, Massachusetts). The materials of the external fixator and Schanz screw (Kangli Medical Group, Jiangsu, China), limited internal fixation screws, and plate (Synthes GmbH, Zuchwil, Switzerland) were scanned and transferred to the computed data to be used directly in simulating the fixation of the fracture models after smoothing.

Fig. 1

The CAD model of the EFLIF for Sanders type 2 calcaneal fractures.

Fig. 2

The CAD model of the plate for Sanders type 2 calcaneal fractures.

Volume mesh generation

Finite element models were constructed using linear tetrahedrons. In this study, there were 765 848 elements with 163 617 nodes in the EFLIF model after meshing, and 1 323 960 elements with 254 345 nodes in the plate model.

Assignment of material properties

The related material properties of a calcaneal fracture finite element model were set up to establish a 3D finite element model similar to the practical model in material parameters and mechanical behaviour. The ligaments were set up as linear elastic materials sustaining tensile stress only. The material parameters of the fixation devices were obtained from Gefen et al⁷ and manufacturers (Table I).

Constraint and loading

Simkin and Stokes⁸ found that the Achilles tendon stress sustains about 50% of the force of the foot. The volunteer’s weight in this study was 64 kg. While standing stationary, one foot sustained 320 N and the Achilles tendon sustained 160 N of pressure. Therefore, 160 N was loaded in the direction of the Achilles tendon to simulate the maximum tensile stress that the calcaneus sustained during non-weight-bearing of the lower limb. The relative displacement and stress distribution were calculated to assess the stability of the two fixation systems. The models were implemented into the Abaqus finite element software (Dassault Systemes). When 160 N was loaded, the tibia, fibula, cuboid, wedge and metatarsal bone were fixed in space.

Clinical data

A total of 50 consecutive patients with closed Sanders type 2 calcaneal fractures were included in this study from January 2010 to December 2013, and divided into two groups according to the treatment (as chosen by the patient): the EFLIF group and the ORIF group. The trial had been approved by the hospital’s ethic committee, and all patients signed informed consent. The inclusion criteria were as follows: patients with unilateral Sanders type 2 calcaneal fractures, diagnosed by radiograph and CT images; ages from 18 to 58 years; and patients with a clear trauma history, and without any other treatment. The exclusion criteria were as follows: age less than 18 years or more than 58 years; Sanders types 1, 3, and 4 calcaneal fractures; bilateral calcaneal fractures or combined with other damage; open or pathological fracture, and infection near the heel combined with serious medical diseases, cannot tolerate surgery or actively co-operate with the post-operative function of exercisers.

Surgical procedure

All patients were positioned supine with continuous epidural anaesthesia. In the EFLIF group, the use of a tourniquet was not preferred. Two 5 mL syringe needles were used to position the medial side of the non-fractured area of the calcaneal tuberosity. Two Schanz screws were then inserted through a small incision to replace the syringe needle after the most appropriate location was verified by a lateral radiographic view. A Schanz screw was inserted into the tibia and navicular bones in the same way. The fixator rods and clamps were assembled to allow for a full visualisation of the subtalar joint on a lateral radiographic view. The clamps were then locked temporarily under traction. A 2.5 cm incision was made under the lateral malleolus to expose the surface of the subtalar joint. The fracture could be reduced under direct vision. Two cannulated screws were inserted to fix the fractures. Finally, the clamps were locked permanently after the reduction was confirmed by fluoroscopy.

In the ORIF group, the patients underwent a standard procedure as described by Benirschke and Sangeorzan.⁹ A thigh tourniquet was used and inflated to 350 mmHg. The fracture was fixed with a calcaneal plate and screws. Finally, the wound was closed in layers with a drain under the flap.

Three months after healing, two examiners assessed pain, walking ability, gait, ankle activity, ankle stability, and alignment, respectively. Hospital stay, blood loss, operation time, intra-operative fluoroscopy times, subtalar joint range of movement, return to previous work, patient satisfaction, and subtalar arthritis were recorded and compared between the two groups.

Statistical analysis

Means (and standard deviations) were calculated for each variable of interest, and t-tests were used to compare the treatment group. The Fisher’s exact test was used to compare the complications between groups, as well as gender and type of accident. A p value of < 0.05 was considered significant.

Finite element analysis: displacement of fractures

In a neutral loading of 160 N, the relative displacement was as follows: 0.1363 mm for the EFLIF in treating a Sanders type 2 calcaneal fracture; 0.1808 mm for the plates and less than 1 mm for both fixation systems.¹⁰

Finite element analysis: von Mises stress on fixation

There was no fixation failure in either fixation system. The highest value of von Mises stress of the external fixator was 30.88 MPa on the Schanz screw (Fig. 3). The highest value of the stress on the screws was 21.16 MPa in the fracture area (Fig. 4). The highest stress value of the plate was 141.9 MPa on the connection of the screws and the plate (Fig. 5). The stress value of the plate was 33% higher than the stress value of the EFLIF.

Fig. 3

External fixator stress nephogram.

Fig. 4

Screw stress nephogram.

Fig. 5

Plate stress nephogram.

Clinical outcome

Age, gender, type of accident, accompanying injuries and comorbidities were similar for the two groups. Follow-up time in the ORIF group was 23.66 months and 25.03 months in EFLIF group, and no significant differences were observed between the two groups (t = 1.686, p = 0.096). The external fixator and the plate were removed at the last follow-up. The timing of surgery, operation time, blood loss, and hospital stay were significantly less in the EFLIF group (p < 0.05). The intra-operative fluoroscopy times were greater in the EFLIF group (4.43 to 6.75, t = 8.070, p = 0.000). The Böhler angle was reconstructed anatomically (Fig. 6); a significant decrease in the Böhler angle was not encountered during the follow-up period.¹¹

Fig. 6

A patient with Sanders type 2 calcaneal fractures treated with EFLIF. From left to right: a) pre-operatively in the lateral view; b) CT pre-operatively; one week post-operatively in the c) lateral and d) axial view; and e) one year post-operatively in the lateral view.

No significant differences were noted in the Böhler angle between the two groups pre-operatively, post-operatively, or at last follow-up (p > 0.05); nor were any significant differences found in the American Orthopedic Foot and Ankle Society (AOFAS) ankle hindfoot scale (83.83 to 84.40, t = 0.994, p = 0.324) or the Visual Analogue Scale (1.63 to 1.56, t = 0.389, p = 0.698) between the two groups. The subtalar joint range of movement, return to previous work, patient satisfaction, and subtalar arthritis were similar between the two groups (Table II).

One external fixation pin tract infection (2.85%) occurred in the EFLIF group, while six superficial infections and one deep infection (22.22%) occurred in the ORIF group, complications in the EFLIF group was significantly lower than ORIF group (p = 0.028, chi-squared). Administration of oral antibiotics and dressing changes were performed to treat the external fixation by pin tract and superficial infections. Implant removal, operative debridement and administration of intravenous antibiotics were performed to treat the deep infection at the sixth week post-operatively.