At one time, pilon fractures were considered nearly impossible to surgically repair. Ruoff and Snider (11) recommended traction as the best treatment option. They were treated with casts, calcaneal pin traction, and pins in plaster technique. Cast management is still appropriate for nondisplaced pilon fractures; however, they are uncommon. Both calcaneal pin traction and pins in plaster have been eclipsed by the use of external fixation. In 1969, Ruedi and Allgower (5) in their landmark work reported on a series of 84 consecutive pilon fractures treated by the principles of the Swiss Study Group/AO. These principles include the following:

This was the first revolution in the treatment of pilon fractures. Seventy-eight fractures were evaluated with an average follow-up of 50.3 months. Seventy-four percent of the patients obtained a good to excellent functional result. Only 9% of patients underwent ankle arthrodesis, and 5% required corrective osteotomy. Seventy-five percent of the fractures were caused by skiing accidents, and only 33% were type III injuries. In a second study, Ruedi (12) reviewed 54 of the original 84 patients 9 years after surgery. He noted that 68% of the results had not changed, whereas 22% had a general improvement. Only in 10% of the patients was there an aggravation of their symptoms.

The important work of Ruedi and Allgower led to an enthusiasm for open reduction with internal fixation (ORIF) (5,12). This is because their conclusion indicated that pilon fractures could be reconstructed and that anatomic reduction of the ankle was necessary to obtain a good result. They also noted that traumatic arthritis would normally develop within 1 year if it was to occur. In addition, some patients could actually improve with time, and patients that demonstrated a good result after 1 year would usually continue to do well.

However, this study later came into question for the high percentage of good results. It was noted that the patient population
was fairly homogeneous, being younger, athletic individuals who sustained lower-velocity-type fractures. To counter this criticism, Ruedi and Allgower (13) reported another series of 75 patients. Forty-seven percent of the patients sustained their fracture during sporting events, 34% had road or work accidents, and 19% had injuries occurring at home. The open fracture rate was 4%. Classification of the fractures demonstrated 25% were type I, 28% type II, and 47% type III. Subjective results from the patients were 80% good, whereas the author’s objective criteria showed 69.4% good results. Only 8% of patients had wound healing problems, and there were no cases of osteomyelitis. The arthrodesis rate was 5.3%. The slightly lower number of good results was attributed to the differences in average patient age, mechanisms of injury, and the experience of the surgeon.

In 1977, Heim and Nasser (14) reviewed a series of 128 patients with 90% good functional results. Heim was the principal surgeon, and 90% of the injuries involved skiing. In 1979, Kellam and Waddell (7) reported their results in 26 cases of pilon fractures. Surgical treatment of a shearing, rotationaltype fracture without major articular impaction yielded 84% good results. However, fractures of the axial compression type with impaction and comminution achieved good results in 53% of patients. From these statistics, it was concluded that the type of fracture had a significant impact on the final results. They also confirmed previous reports that traumatic arthritis would occur within 1 year in patients destined for this sequelae (5,12). The incidence of subsequent arthrosis was directly related to the degree of articular damage and the congruity of the articular surface. It was speculated that avascular necrosis of the smaller, comminuted fragments of subchondral bone readily led to progressive deformity accelerating the arthritic process (7). Interestingly, Resch et al (15) reported, in 1986, if there was poor joint surface reconstruction after surgery, the incidence of late arthritis was higher than in patients managed nonsurgically.

After the articles by Ruedi and Allgower, ORIF became the mainstay of pilon fracture management (5,7,12,13 and 14). Type I fractures (nondisplaced) were normally managed with cast immobilization but are rare injuries. Type II fractures lent themselves well to ORIF with reasonable expectations of obtaining an anatomic reconstruction. However, the type III fractures remained problematic. Failure to achieve an anatomic reduction with stable fixation would result in an increased risk of complications in
an already highly traumatized extremity. A critical decision was the appropriate timing for a surgical procedure. It was thought surgery should be performed as soon after the injury as possible. This usually meant within 8 to 12 hours before the onset of significant edema and the development of fracture blisters. If surgery was to be delayed, then one should consider waiting 7 to 10 days for the local skin condition to stabilize.

If surgery was delayed, the patient was evaluated for possible compartment syndrome. Fracture blister formation could occur and represents a separation of the epidermis and dermal layers. They can be filled with a serous or hemorrhagic fluid. Hemorrhagic fracture blisters indicate a more significant soft tissue injury. The fluid is sterile inside a fracture blister unless it is ruptured (16). The best advice was not to operate through fracture blisters. Small, unruptured, serous filled blisters away from the incisional site did not necessarily preclude surgery. However, hemorrhagic fracture blisters were a contraindication to surgery. Varella et al (17) reported a 60% incidence of infection when surgery was performed through open fracture blisters. Fracture blisters can be allowed to resolve by themselves or be débrided. Personal preference is débridement followed with wet to dry dressings and periodic air drying. With a pilon fracture awaiting surgery, this was best accomplished by placing the patient in a cast followed by windowing to allow wound care. The cast offered some stabity, which decreased the pain and prevented unnecessary movement of the soft tissue and osseous structures. If the patient was in an external fixator, this process is made easier.

The traditional AO approach for pilon fractures was two incisions assuming both the fibula and tibia are fractured. The fibula was approached by a posterolateral incision and the tibia by an anteromedial incision. Both the fibular and tibial fractures were exposed prior to definitive reduction. The posterolateral incision allowed access to the fibula, which was directly reduced and fixated usually with a combination of screws and plate. Fibular reconstruction usually preceded that of the tibia. The large tibial incision started 1 cm lateral to the crest of the distal tibial shaft and then curved over the anteromedial aspect of the medial malleolus. The important point was to maintain a 7-cm skin bridge between the two incisions to prevent necrosis. If only the tibia was fractured, then the incision was placed anteriorly, anterolaterally, anteromedially, medially, or posteromedially depending on the fracture configuration. Sometimes, even an isolated tibial fracture might require a two-incisional approach. Incision placement was along a major fracture line and guided by the CT scan. The dissection is carried to bone without raising soft tissue flaps, following the planes created by the fracture avoiding further devitalization of the tissue. The fracture fragments were reduced under direct visualization using the talus as a template. An effort was made to reduce every fragment anatomically. Bone fragments were provisionally fixated with guide pins, Kirschner wires (K-wires), and reduction forceps. A cancellous bone graft was placed in the metaphyseal defect. To rigidly fixate the fracture and to prevent late varus deformity, a buttress plate was usually applied to the medial aspect of the tibia. The plate could also be applied to the anterolateral, anterior, or, rarely, the posterior aspect of the tibia depending on the fracture configuration. There were several different styles of plates that could be utilized, such as the cloverleaf, spoon, or limited contact dynamic compression plate (LCDCP). Lag screws could be placed through or outside the plate to obtain interfragmental compression of the fracture fragments. After the plate was applied to the tibia, final intraoperative radiographs were obtained to assess if the reduction was anatomic and placement of the internal fixation appropriate (Fig. 110.3).

The second revolution in pilon fracture management was external fixation with or without limited ORIF (Fig. 110.4). This technique was introduced by Leach (18) in 1964. Scheck (19), in 1965, using the same technique reported satisfactory results in five cases of pilon fractures in which four of the injuries were open. External fixation with or without internal fixation became popular because of experiences with traditional ORIF that resulted in significant decrease in functional results and an increased rate of complications in pilon fractures, especially the type III (20,21,22 and 23). In a series of 19 patients, Pierce and Heinrich (20) reported poor results despite treatment by varied methods. Ovadia and Beals (9) reviewed 145 pilon fractures, 80 of which were treated with traditional ORIF. They developed their own classification system and reported that the fracture type was extremely important in determining the outcome. Their type I, II, and III fractures had a 70% good to excellent result. However, their type IV and V fractures representing 63 injuries, with 34 undergoing ORIF, demonstrated 47% good to excellent results. The infection rate was 6%, and the arthrodesis rate 12%. In 1986, Dillin and Slaubaugh (21) reported on eight patients with severe pilon fractures who underwent ORIF. In this group, 100% developed either wound breakdown, chronic drainage, or osteomyelitis. All eight patients required additional procedures.

In 1983, Bourne et al (22) reported on 42 pilon fractures, nine of which were open injuries. In 27% of their cases, other injuries were present. Two patients had neurovascular compromise. The fractures were categorized in the Ruedi and Allgower system as type I (26%), type II (29%), and type III (45%). The results were reported as type I 86% good to fair, type II 80% good to fair, and type III 44% good to fair. In the type III fractures, the nonunion rate was 25%, the infection rate 13%, and the arthrodesis rate 32%. Arthrodesis was performed in seven fractures (17%) with six of the patients having type III injuries. Bourne (3), in 1989, reported again on the 42 pilon fractures with a follow-up at a mean of 53 months after treatment. He noted that type I fractures were usually caused by a torsion mechanism, whereas the type II and III fractures were associated with high-velocity injuries. The results indicated that type I and type II fractures were amenable to anatomic reduction and stable fixation with an 80% success rate. However, the type III fractures presented a more difficult problem and had a satisfactory outcome in only 32% of patients.