Minimally invasive (sinus tarsi) approach for calcaneal fractures

This study was approved by the Institutional Review Board of the First Affiliated Hospital of Dalian Medical University, and all patients provided written informed consent for the surgery and for this study. We retrospectively reviewed the clinical data of patients with calcaneal fractures who were treated by using a minimally invasive sinus tarsi approach in the First Affiliated Hospital of Dalian Medical University between March 2014 and October 2015. The eligible patients met the following inclusion criteria: (1) age >18 years; (2) with closed fracture of the calcaneus; (3) with Sanders types II or III fractures [1]; (4) with fractures fixed by using our newly self-designed combined locking plate; and (5) with a minimum of 6 months of clinical follow-up. Patients with history of calcaneal, ankle, or other foot fractures were excluded.

The plate (GJPS(I)67; 1.5 mm in thickness and 13 mm in width; made of titanium alloy owing to the better biocompatibility, flexibility, and a lower resonance of this material [8]; Puwei Medical Instrument Co., Ltd, Shanghai, China; patent no. ZL 2014 2 0269654.X) consisted of two sheets, mutually independent fixed arms (forearm, 42 mm in length; rear arm, 39 mm in length), a coupling screw, and six locking screws (Fig. 1). There are three screw holes in the forearm/rear arm to lockingly fix the anterior process of the calcaneus, calcaneal body, and bones behind the calcaneal body, respectively. The head end of the forearm/rear arm is of streamlined shape, whereas the tail end can be linked by the coupling screw through their chimeric structure.

The biomechanical characteristics of our self-designed combined plate were tested by Biomechanics Laboratory of Changzhou Waston Medical Appliance Company (Changzhou, China) by using the standard electro-mechanical testing machine (Instron ElectroPuls E10000; Instron Systems, Norwood, MA, USA) equipped with a 10-kN Instron load cell. The vertical load-bearing capacity of the combined plate was tested under static (in which the loading rate was 2 mm/min and the maximum load was recorded by monitoring the force-displacement curves) and dynamic (performed in load-controlled mode with sinusoidal wave at a frequency of 5 Hz and maximum load of 500 N) conditions (Fig. 2a, b). For the resistance to bending loads, the force-displacement curve was also monitored and the load to failure was recorded at the loading velocity of 2 mm/min and with the span of 30 mm (Fig. 2c, d).

All procedures were performed by the same surgeon (Zhe Wang) as previously described [6, 7, 9]. After admission, the affected feet of the patients were fixed by using a plaster or provisional brace and elevated to prevent the exacerbation of injuries in the soft tissues that can be induced by activity. For patients with slight damages to the soft tissue, the surgery was scheduled within 48–72 h; however, for patients with severe tissue damages (accompanied by obvious swelling and tension blisters), the surgery was arranged after 7–10 days to confirm the reduction of swelling, improvement of soft tissue condition, and a positive wrinkle test. This led to a mean interval from injury to surgery of 4.7 ± 2.4 days (range, 2–9 days).

Surgery was performed with the patients placed in a lateral decubitus position under epidural or subarachnoid anaesthesia (Additional file 1). An electric pneumatic tourniquet (pressure, 45 KPa) was used on the affected limb to reduce blood loss and improve the operative field visually. One or two Steinmann pins were first drilled from the calcaneus to the head of the talus to correct the Böhler angle and restore the Gissane angle, which was confirmed under C-arm fluoroscopy. Then, a 3–4-cm incision was made over the sinus tarsi (beginning from the tip of the lateral malleolus to the proximal cuboid) to expose the lateral cortex of the anterior process of the calcaneus (carefully, to avoid injuring the lateral sural cutaneous nerve, peroneus longus, and brevis tendon) (Fig. 3a, b). Subsequently, a subcutaneous tunnel was created by using a small periosteal elevator in the above incision, and our self-designed plate was inserted to the corresponding position of the lateral wall of the calcaneus, which was temporarily held in place with two Kirschner wires (Fig. 3c, d). After a satisfactory reduction was achieved under C-arm fluoroscopy, six locking screws and a coupling screw were screwed followed by the removal of the Steinmann pin and closure of the incision in layers.

Postoperatively, prophylactic antibiotics were given to prevent surgical site infection and plaster immobilization was performed to protect the wound. The patients were encouraged to do toe flexion and dorsiflexion exercises 24 h after the operation. Stitches were removed after 2 weeks of surgery, and weight bearing was gradually advanced to regain full range of motion and strength at 4 weeks after surgery, during which the plaster cast was removed.

The operative data were recorded, including operative time (min; measured from the cut to the suture of the incision), hospital stay (days), cumulative intraoperative radiation time (min; provided by the fluoroscopic apparatus), fracture healing time (days), and the incidence of complications. Fracture healing was defined as the bridging of the trabecular bone and the disappearance of the trabecular fracture line on the radiograph. The lateral and axial radiographs were also obtained preoperatively, postoperatively, and at the last follow-up to judge the reduction of the calcaneus, including the Böhler angle (i.e. the angle between the line drawn from the highest point of the anterior process to the highest point of the posterior facet and the line tangent to the tuberosity); Gissane angle (i.e. the angle between the line tangent to the articular surface of the medial posterior facet and the line of the anterior end of the posterior facet to the dorsal edge of the calcaneocuboid facet); calcaneal height (i.e. the perpendicular distance from the inferior cortex of the calcaneus to the top of the medial posterior facet); calcaneal width (i.e. the distance between the external cortex of the medial malleolus and the lateral cortex of the most lateral calcaneal fracture fragment); and calcaneal length (i.e. the orthogonal distance from the most posterior aspect of the calcaneus to the most distal edge of the calcaneocuboid joint) [10, 11]. The clinical functional outcomes were assessed by using the Maryland foot score (MFS), which is a 100-point scoring system with 40 points for function, 45 points for pain, 10 points for cosmesis, and 5 points for movement of the ankle, subtalar, midfoot, and metatarsophalangeal joints. The scores were defined as follows: 90–100 excellent, 75–89 good, 50–74 fair, and <50 poor [12]. All data collection and measurement procedures were performed by Zhe Wang and checked by Sheng Long Li. Measurement of reduction parameters was accomplished with the DWRuler software.

All data are expressed as n (%) or mean ± standard deviation (SD) and analysed by using SPSS 18.0 statistical software (SPSS Inc., Chicago, IL, USA). The preoperative and postoperative radiological results were compared by using paired Fisher’s exact test (because of an expected frequency of <5), and an independent t test was used to compare the functional outcomes between the different Sanders classifications. A value of p < 0.05 was considered to indicate a significant difference.

From March 2014 to October 2015, 18 patients (14 men, 4 women) with calcaneal fractures were enrolled in this study. Their average age at the time of the calcaneal fracture was 50.4 ± 10.1 years (range, 32–66 years). The mechanism of injury was falling from a height in 14 cases, traffic accidents in 2 cases, strike injury in 1 case, and others in 1 case. All fractures were unilateral, with the left side involved in 7 cases and the right side in 11 cases. In addition, the lumbar vertebral, clavicular, and ankle fractures were complicated in 1 case. Six patients had a >5 years smoking history and one patient had a history of excessive consumption of alcohol, and they were advised to stop smoking and abstain from alcohol until wound healing; on the other hand, two patients had diabetes, and their blood glucose was controlled to prevent complications [13, 14]. According to the Sanders CT scan classification system, there were 14 feet with type II fractures (8 type IIA, 4 type IIB, and 2 type IIC) and 4 feet with type III fractures (2 type IIIAB, 1 type IIIAC, and 1 type IIIBC) (Table 1, Additional file 2: Table S1) [1].

The analysis of biomechanical properties demonstrated that the plate could bear 1396.03 N in maximum vertical load-bearing capacity and 427.15 N in maximum resistance to bending loads (Fig. 4). Under a cyclic loading of 500 N, plate breakage was observed after 93,003 cycles.

The surgery was successful in all patients, with a mean operative time of 64.4 ± 8.0 min (range, 50–80 min), mean radiation time of 13.0 ± 3.5 s (range, 8–21 s), and mean hospital stay of 8.9 ± 2.2 days (range, 5–13 days). The radiographs showed that all fractures healed uneventfully, with a mean time to bone union of 89.3 ± 11.0 days (range, 75–113 days). The radiographs also demonstrated adequate reduction of the fracture (Table 2, Additional file 2: Table S1), with the calcaneal width, length, height, Böhler angle, and Gissane angle significantly corrected from preoperatively to 3 months postoperatively and the last follow-up (range, 6–13.5 months; mean, 9.3 ± 3.7 months). However, there were no significant differences when comparing the calcaneal width, length, height, Böhler angle, and Gissane angle between 3 months after the operation and at the last follow-up, indicating that this reduction effect can be maintained but not lost during follow-up. The overall mean MFS was 88.1 ± 8.8, in which excellent outcomes were achieved in 11 patients, good in 4, and fair in 3 (excellent and good rate, 83.3% (15 of 18)). Further comparison between type II and type III fractures showed no statistical significances in the mean MFS (88.1 ± 8.8 vs 87.8 ± 10.1, p = 0.9) and in the excellent and good rate (85.7 vs 75.0%, p = 1.0), suggesting that our self-designed plate may be equally effective for these two types of fractures (Table 3, Additional file 2: Table S1). Moreover, no screw loosening and screw or plate breakage was observed, and none of the patients developed incision infection, poor wound healing, and other complications during the follow-up, proving the favourable safety profile of the plate. Radiographs of three typical cases (case 1, Fig. 5a–f; case 2, Fig. 6a–h; case 12, Fig. 7a–h) were provided to illustrate the preoperative and postoperative changes.