Our main objective was to observe the convenience of the fibular notch approach in the treatment of complex ankle fractures and its effect on fracture reduction and efficacy. It should be noted that the cases in the current study included both classical pilon fractures (AO/OTA 43-B and 43-C) and special ankle fractures (AO/OTA 44-C), although strictly speaking, it might not be appropriate to discuss the fractures as a whole. Sometimes ankle fractures involving the distal tibial plafond are caused by a combined injury mechanism, which typically leads to a confusing classification of the fracture, such as a "logspatter" injury  versus an atypical pilon fracture, as well as a posterior malleolus versus a "posterior pilon" fracture . We would like to use the words "fractures involving the distal tibial plafond" rather than "tibia pilon fractures" to describe the confusing ankle fractures in the present study.
According to the injury mechanism, when a large axial load is combined with valgus violence, the talar dome has a tendency to drive into the lateral portion of the plafond, resulting in a compression fracture of the plafond, occasionally with comminuted articular fragments. In some patients, the impaction between the talus and fibula can lead to concomitant injury of the fibula and distal tibiofibula. There are two points of concern in the treatment of these complex fractures. One is to try to reconstruct the anatomic bony structure, including the articular surface and metaphysis, to prevent traumatic arthritis, and the other is to make a safe incision on the traumatized and thin surrounding soft tissue to avoid soft tissue complications [9, 18]. Clearly, the ideal surgical approach is designed for easy exposure and for mechanically appropriate fixation without increased soft tissue complications . However, for severe distal tibia fractures involving the plafond, a single incision bears the defect of inadequate exposure of the joint, while double incisions not in the diagonal axis of the ankle sometimes fail to meet the requirement of a distance of 7 cm between incisions [19, 20].
In fact, few other fractures have been designed with so many surgical incisions in such a narrow area as pilon fractures. A variety of approaches, including anterolateral , posterolateral , and lateral approaches , have been well described. We acknowledge that most fractures in the distal tibia can be treated through traditional approaches; however, these approaches do not seem to be a solution for all fractures, especially when a severely comminuted intra-articular fracture is present . Because of the obstruction of the distal fibula and the talus, it is actually difficult to obtain sufficient operative visualization of the die-punch fragments through a traditional single incision, while multiple approaches for direct ORIF may result in a high risk of soft tissue complications . As a result, with the absence of intraoperative CT to help examine reduction quality, poor restoration of the articular surface is not uncommon, even in patients treated by skilled surgeons .
Since Adams JC described the transfibular approach for ankle arthrodesis in 1948, the transfibular approach has been widely used by Napiontek M , Mann RA et al , with the best advantage of fully exposing the articular surface of the ankle. Inspired by this approach, we tried to apply a novel approach similar to it, the fibular notch approach introduced in the present study, for the treatment of some specific fractures. An epidemiological study showed that syndesmotic injury can occur in up to 20% of classical ankle fractures  and 15% of pilon fractures . This provides surgeons an opportunity to utilize the injured lateral structure to fully expose the most difficult but critical articular surface without significantly increasing iatrogenic injury. Nabil A  described the anatomic characteristics of the distal fibular notch on CT scans. At the location of the axial cuts 1 cm proximal to the tibiotalar joint line, which was also the exact plane of clinical significance in our present patients, the length of the anterior facet of the notch was 11.2 ± 1.9 mm and that of posterior facet was 14.89 ± 2.72 mm and thus 36 mm in total. This means that for the distal tibia, which is approximately 45 mm in diameter at this location, two-thirds of the fibular notch is blocked by the lateral malleolus, and it is the exact main factor blocking the exposure of the articular surface in the widely used anterolateral (AL) and posterolateral (PL) approaches.
Anatomically, the distal tibia and fibula are tightly linked by four syndesmotic ligaments, and the distal fibula is further enhanced by the lateral ankle ligament complex . While using the fibular notch approach, the ATFL is the only possible ligament that sometimes needs to be cut off for sufficient exposure of the surgical field. The other ligaments, including the PITFL, the CFL and the PTFL, which extend obliquely from anterior-lateral to posterior-medial, are very unlikely to be involved in the original injury (less than 10% involved) [29, 30] and will not be obstacles to the intraoperative lateral rotation of the distal fibula. Typically, the AL approach is widely used in fractures involving the anterolateral components of the distal tibia with good visualization, and the PL approach is more advantageous for the treatment of posterior column and Volkmann fragments. The fibular notch approach is a combination of the two and has both advantages. As the fibula is rotated posteriorly, an anterior interval similar to but broader than that in the AL approach can be obtained. With the first temporary reduction of the fibula, the same posterior interval as is achieved in the traditional PL approach can be achieved. Of particular significance, it is convenient to retract the distal fibula again to open the syndesmosis to examine the articular fragments and the articular surface in suspicious patients without blocked visualization due to the reduced and temporarily fixed fragments at the articular margins . The posterior-anterior-posterior process of retracting the distal fibula and the “second look” technique have the advantage of avoiding the embarrassment caused by uneven articular surfaces on postoperative CT scans. It should be noted that the two intervals in the fibular notch approach are not equally effective. The posterior interval is essentially designed to assist the anterior interval for facilitating fixation from the posterior to the anterior direction. Therefore, although the incision is located at the posterior border of the fibula rather than at the traditional midway point between the Achilles tendon and lateral malleolus , it is not difficult to complete manipulation behind the ankle.
To our knowledge, four similar incisions or approaches have been described in prior articles. The first is the transfibular approach used for ankle arthrodesis . It is also a real approach that completely exposes the fibular notch, and the distal fibula is completely or partially amputated rather than temporarily removed. The second is the trans-fibular-fracture approach described by Gonzalez TA for fixation of ankle fractures . In this approach, all manipulations of distal tibial exploration and reduction are performed through the fracture line of the fibula. It is not suitable for low fibula tip fractures (Weber A ankle fractures) or high-level transverse fibular fractures, and the fibular notch is not exposed. The third is an oblique incision that was introduced by Niall P et al  in 2016 for simultaneous open reduction and internal fixation of the posterior malleolus and anterior syndesmosis. The authors fixed posterior malleolar fractures via the window between the peroneal tendons and the posterior aspect of the fibula and explored anterior syndesmosis via dissection anterior to the fibula. The authors also did not mention the exposure of the fibular notch. The last is a case report published in January 2019 during the follow-up of our case . The authors used almost the same approach as ours to treat a tibial pilon fracture where only the lateral part of the distal tibia was affected. The incision was made along the anterior border of the fibula, and the distal fibula was retracted posteriorly to explore the lateral aspect of the tibial pilon. Unlike us, the author did not make a posterior interval to address the posterior ankle fracture, nor did they show the posterior-anterior-posterior shift of the distal fibula or utilize the “second look” technique, while these manipulations are of paramount importance for the observation of complex intra-articular fractures. In addition, the authors did not describe the ATFL closely related to the retracted fibula.
The vast majority of patients in the study achieved satisfactory results after a relatively short-term follow-up, without severe PTOA. Except for anatomical reduction of the tibial articular fragments in the present study, accurate reduction of both the fibular fracture and the syndesmosis were considered to be other important factors associated with good functional recovery. The restoration of lateral malleolar length is of paramount importance for optimal efficacy [34, 35]. With the direct visualization of the fibular notch approach, shortening and rotation of the fibula and occult mismatch of the inferior tibiofibular syndesmosis can be avoided effectively. However, four patients with falling injuries still had mild PTOA, which might be related to the original injury of the articular cartilage, and this relatively short follow-up time still needs to be further extended to increase reliability. Overall, a low rate of soft tissue complications was presented in our study, except for one patient with crushed soft tissue who had delayed wound healing and superficial infection, which supported the idea that a single incision had a lower risk of soft tissue complications.
We acknowledge that the fibular notch approach has some drawbacks. It has fewer indications and is not a conventional approach. It can only be considered when comminuted tibial plafond fractures are combined with fibular fractures and distal tibiofibular syndesmosis injuries, that is, specific Weber type C fractures caused by axial load and rotational force are more likely to be associated with the indications. Additionally, except in a few patients in which the ATFL was broken in the original injury, the ATFL was cut off, which may potentially have increased the tendency for postoperative ankle instability. Interestingly, with the soft tissue being repaired carefully, 4 patients in the present study complained of slight ankle stiffness and no one complained of ankle instability. Perhaps it is the same as cutting off the CFL in an extensive lateral approach to a calcaneal fracture, which is not prone to postoperative instability due to soft tissue contracture and scarring.