Open pilon fractures of the distal end of the tibia typically occur as the result of a high-energy mechanism. Axial load and shear forces often lead to a severe soft-tissue injury and a comminuted fracture pattern. The severity of injury to the soft-tissue envelope, the fracture pattern, the patient profile, and the surgeon's expertise are factors that contribute to the choice of treatment method and affect the prognosis. Open reduction and internal fixation most reliably ensure acceptable restoration of length, rotation, and alignment of the metadiaphyseal segment with anatomic restoration of the articular surface; however, the potential for soft-tissue complications has led many surgeons to choose external fixation with or without limited internal fixation as definitive treatment1-4.
Historically, prior to the 1979 landmark study by Rüedi and Allgöwer5, open treatment of pilon fractures was uncommon. In response to the predominant recommendations for nonoperative treatment6,7, Rüedi and Allgöwer compared open reduction and internal fixation with nonoperative treatment. They noted improved results in the operatively treated group and advocated routine open reduction and internal fixation in the treatment of displaced fractures. A subsequent critical analysis of their study population revealed that the mechanism of injury for most fractures was a twist and a fall while skiing8. Compared with the fractures sustained in motor vehicle crashes or falls from a height, these patients typically sustain a less severe soft-tissue injury8. When the treatment principles of Rüedi and Allgöwer were extrapolated to high-energy pilon fractures with associated soft-tissue injury and comminution, comparable results were not obtained9-14. In fact, the prevalence of complications, including infection, wound dehiscence, skin slough, nonunion, malunion, and posttraumatic arthritis, was markedly increased9,15. These findings led many surgeons to utilize various forms of external fixation with or without limited internal fixation as a first choice2,14,16. Although the risk of infection and soft-tissue complications may be decreased with the use of external fixation, this modality may not allow optimal reduction of the joint surface and thus jeopardizes the long-term outcome. There remains a paucity of data pertaining to the prevalence of acute complications and long-term outcomes following treatment of open pilon fractures14,17-22.
We sought to evaluate outcomes following open reduction and internal fixation of high-energy open pilon fractures. Outcomes were assessed retrospectively in a series of fifty-nine patients who underwent treatment under the direction of a single surgeon during an eleven-year period.
Between December 1995 and October 2007, 186 pilon fractures in patients presenting to a single level-I trauma center were managed operatively under the direction of the senior author (D.E.A.). No fracture was managed nonoperatively during this time period. Of the 186 pilon fractures, sixty-eight were open. A prospectively maintained database was available for retrospective analysis. Fifty-nine patients were followed for a minimum of one year, with fifty-four patients being followed for two years or more. The mean duration of follow-up was thirty-four months (range, one to ten years). All patients were included for the assessment of wound and fracture-healing. Patients who had less than two years of follow-up were excluded from functional outcome assessment. Functional outcome assessment was performed on thirty-eight patients at two years after the injury or later.
Routine follow-up was performed at one, three, six, twelve, twenty-six, and fifty-two weeks, with an annual evaluation thereafter. Assessment included a subjective evaluation of pain, evaluation of medication use, and a physical examination to evaluate the soft tissues and range of motion. Radiographs were made to evaluate the quality of fracture reduction, the stability of fixation, fracture alignment, and union. Clinical and radiographic data were obtained from a chart review. Functional outcome was assessed for patients with follow-up of two years or more with use of the Rand 36-Item Health Survey (Rand-36) measure of health-related quality of life23 (Short Form-36 Version 2.0; SF-36v2). Ankle-specific outcome was assessed with use of the modified Mazur scoring system24. The institutional review boards of Westchester Medical Center and New York Medical College approved the study design.
Treatment Regimen
Initial Management
Initial treatment in the emergency department included closed reduction, placement of a dry sterile dressing at the wound site, and application of a long leg splint. All patients were seen and evaluated by the general surgical trauma service with use of the standard American College of Surgeons Advanced Trauma Life Support protocol25. Patients with an open fracture of grade I, II, or IIIA, according to the system of Gustilo and Anderson26,27, were treated with intravenous administration of a first-generation cephalosporin. Patients with an open fracture injury of grade IIIB or IIIC also received an intravenous aminoglycoside. Any open injury was addressed in the operative suite as soon as possible following general surgical stabilization. Devitalized bone and soft tissue were débrided. Irrigation was performed with a minimum of 9 L of saline solution with use of pulsatile lavage. Wounds were loosely reapproximated or remained open following the initial débridement. Nine patients underwent definitive open reduction and internal fixation at the time of the initial irrigation and débridement, and fifty patients underwent temporizing external fixation to restore length, rotation, and alignment and to protect the overlying soft tissues while awaiting a definitive procedure. In all patients, components from either the AO/ASIF external fixation system (Synthes, Paoli, Pennsylvania) or Hoffmann-II external fixation system (Howmedica Osteonics, Mahwah, New Jersey) were applied in a delta configuration. Schanz pins were placed outside the zone of injury and expected surgical field. Tibial Schanz pins were placed from anterior to posterior by means of stab wounds after predrilling. A centrally threaded, calcaneal transfixion pin was placed from medial to lateral.
Serial débridement was performed every forty-eight to seventy-two hours until all obviously devitalized tissue and foreign bodies were removed. Early in the series, wet-to-dry dressing changes were used for the treatment of some open wounds pending delayed primary closure or flap coverage. In later years, vacuum-assisted suction drains (V.A.C.; KCI, San Antonio, Texas) were used prior to delayed primary closure or flap coverage. Subsequent to delayed primary closure and pending the definitive procedure, dry sterile dressing changes were performed approximately every second or third day. Antibiotics were continued for forty-eight hours after it was determined that the wound was stable. The wound was deemed stable when it was amenable to primary closure or coverage by means of a free tissue transfer. Early in the series, some patients underwent open reduction and internal fixation of the fibula at the time of the initial débridement. This practice was abandoned later in the series because the lateral incision sometimes precluded the desired definitive approach or because nonanatomic fixation of the fibula compromised the subsequent procedure.
Definitive Management
Following initial surgical management, a computed tomography scan was made to aid with preoperative planning of the definitive procedure. The operative plan was developed on the basis of the fracture configuration and the status of the soft tissues. The external fixator was sometimes retained and used as a reduction aid during the definitive procedure. In some patients, the external fixator was removed and a femoral distractor was used during the definitive procedure. The decision to remove or use the external fixator was based on the quality of the reduction, the effect of retaining the fixator on the surgical approach, and the positioning of the hardware. In thirty-seven patients, open reduction and internal fixation was carried out with use of a lateral surgical approach28. Alternative and supplemental approaches were used as indicated by the fracture pattern and soft-tissue injury. The direct anterior approach was the most commonly chosen alternative approach. An attempt was made to incorporate traumatic wounds, particularly when subsequent flap coverage was anticipated. Relatively small implants such as one-third tubular plates were generally applied by means of limited incisions and in an antiglide-type fashion. The small size of these incisions (typically 2 to 4 cm in length) allowed them to be placed in proximity to other wounds. The resulting skin bridges were often of a size that would historically be considered inadequate.
Soft tissues were assessed on an ongoing basis. Plastic surgical consultation was obtained when extensive soft-tissue injury was present and when delayed primary closure of the traumatic wound was not possible. Soft-tissue reconstruction was typically performed on the day of definitive fixation. A total of seventeen free vascularized muscle flaps were performed. No rotational or fasciocutaneous flaps were used. The gracilis muscle was the donor of choice when the defect was small. For large defects, either a rectus or latissimus dorsi free flap was used. All patients were managed with a well-padded, posterior short-leg splint with side gutters at the ankle at the completion of definitive fixation and following soft-tissue reconstruction.
Postoperative Care and Evaluation
Following definitive fixation, patients were maintained at bed rest for two days. The affected extremity was elevated approximately 6 to 8 in (15 to 20 cm) above the horizontal in an attempt to minimize postoperative edema without compromising arterial perfusion.
The splint was changed and the wound assessed on postoperative day 2. If the wound margins appeared tenuous or substantial edema persisted, the patient was maintained at bed rest for an additional twenty-four to forty-eight hours. When the soft tissues permitted, the patient was mobilized without bearing weight on the involved lower extremity. Prophylaxis for deep venous thrombosis was provided with use of sequential compression garments and/or low-molecular-weight heparin throughout the time the patient remained hospitalized and at bed rest. Patients at high risk for deep venous thrombosis, those with a contraindication to alternative modalities, and those with a history of deep venous thrombosis or pulmonary embolus underwent placement of an inferior vena cava filter. Screening for deep venous thrombosis with use of Doppler ultrasonography was generally performed prior to all procedures except the initial surgical procedures. Anticoagulation therapy was discontinued at the time of discharge in the patients who were able to walk with crutches or a walker.
Patients were evaluated by the senior author (D.E.A.) at one, three, six, twelve, twenty-six, and fifty-two weeks and annually thereafter. All patients remained in splint or cast immobilization until the wounds were healed and the sutures could be removed. When the soft tissues allowed, active and active-assisted range-of-motion exercises of the affected ankle and subtalar joint were initiated. Patients were advised to remain non-weight-bearing for a minimum of eight weeks, and then they progressed to full weight-bearing over four to twelve additional weeks as determined by the fracture pattern, the intraoperative assessment of construct stability, and radiographic findings at the time of follow-up.
Routine clinical evaluation included documentation of the patient's subjective report of pain, medication requirements, return to work, ability to perform activities of daily living, and ability to resume recreational activities. General health was assessed with use of the two components of the SF-36v223. The physical component of the SF-36v2 questionnaire incorporates questions pertaining to sports activities, activities of daily living, and lifestyle adjustments necessitated by the injury. The mental component of the SF-36v2 incorporates questions pertaining to emotional adjustment to the physical injury and questions relating to mood and affect. An independent observer conducted thirty-eight telephone interviews of the patients with at least two years of follow-up while following a script provided by the distributor of the SF-36v2 outcome tool.
At each postoperative visit, the physical examination included an assessment of wound-healing, the status of the soft tissues, and the range of motion. Radiographs were assessed for the quality of fracture reduction, stability of fixation, and fracture alignment and union. For the purposes of this study, no attempt was made to quantify the severity of posttraumatic arthritis. Ankle function was assessed with use of the modified Mazur score24. A maximum score of 100 points correlates with the worst possible outcome. Fifty points are directly attributable to pain. Additional points are attributable to the walking distance (8 points), the use of supports or orthoses (8 points), the ability to run (5 points), the ability to do toe raises (5 points), the ability to walk up and down stairs (3 points), the ability to walk up and down a hill (3 points), the presence of a limp (8 points), the presence of swelling (3 points), and the range of motion (7 points).
Statistical Methods
We provide inferences on patient outcomes that can be expected two years after surgery or later. Confidence intervals were calculated for various age groups divided into ten-year intervals with use of standard statistical methods, which assume the outcome to be normally distributed. Average scores on the Rand-36 measure of health-related quality of life (SF-36v2) were directly compared with age-matched United States norms. The physical component and the mental component were assessed individually with use of the Fisher exact test. Multivariate regression analysis was used to determine significant predictors of infection. The presence of infection was designated as an independent variable. Patients with Gustilo grade-IIIB open fractures were compared with patients with Gustilo grade-I, II, and IIIA open fractures for a range of selected outcomes. A p value of =0.05 was considered significant.
Source of Funding
We did not receive any funding for this study.
Demographics and Injury Pattern
Patient demographics and injury details are shown in Table I.
Complications are reported in Table II. Fifty-two of the fifty-nine fractures healed following initial open reduction and internal fixation. On the basis of the radiographic assessment, no patient experienced loss of reduction or loss of fixation. Six patients underwent bone-grafting and limited revision of internal fixation to treat a nonunion or an impending nonunion due to bone loss. In five of these six patients, a bone graft was prospectively planned at the time of the index procedure. All patients receiving a bone graft had sustained an AO/OTA type-C3 injury29, and all of these fractures progressed to union following the bone-grafting procedure. One patient with a grade-IIIB injury required amputation following failed free tissue transfer.
Two patients (3%) had a deep infection diagnosed on the basis of positive intraoperative wound cultures. In one patient, bone-grafting was performed one year after injury to treat a nonunion; at that time, deep wound cultures were positive and the patient was treated for six weeks with culture-specific intravenous antibiotics. Another patient underwent staged reconstruction, including iliac crest bone-grafting, to treat a segmental distal diaphyseal and metaphyseal defect. The fracture healed, the patient underwent removal of symptomatic hardware at two years after the injury, and a postoperative wound infection developed, requiring irrigation and débridement. He was also treated with six weeks of culture-specific intravenous antibiotics. No patient had chronic osteomyelitis develop. Three patients (5%), who presented with superficial erythema but no drainage, were diagnosed as having a superficial wound infection or cellulitis. All three were treated successfully with oral antibiotics. In one patient, partial-thickness necrosis developed in a free vascularized muscle flap, which healed successfully with use of serial wet-to-dry dressing changes.
Fifty-nine patients were followed for a minimum of one year, with fifty-four patients being followed for two years or more. Patients with less than two years of follow-up were excluded from functional outcome assessment. The median duration of follow-up was thirty-four months (range, one to ten years). All patients were included when wound and fracture-healing were assessed. Functional outcome assessment was performed on thirty-eight patients at two years after the injury or later.
The average physical component and mental component scores on the SF-36v2 were 40.3 and 54.9, respectively (Table II). The average physical component scores for the age groups from thirty-five to forty-four years, forty-five to fifty-four years, fifty-five to sixty-four years, sixty-five to seventy-four years, and seventy-five years and older were 43, 41.2, 42, 35.2, and 28.8, respectively. The United States norms of 51.8, 49.3, 47.7, 42.7, and 41 were higher in each age group; however, the difference was significant (p < 0.05) only in patients in the forty-five to fifty-four-year age group (Fig. 1). Study patients scored better than age-matched norms on the mental component of the SF-36v2 in all groups except those from fifty-five to sixty-four years old and those from sixty-five to seventy-four years old; however, the difference was insignificant for all the comparisons (Fig. 1). The average modified Mazur score was 44.8 (Table II). The result was considered excellent in 7%, good in 17%, fair in 14%, and poor in 62% of the patients. Linear regression revealed no association between age (p < 0.38), interval between provisional and definitive fixation (p < 0.8), or grade of open fracture (p < 0.2) and subsequent infection; however, the presence of an AO/OTA type-C3 fracture had a significant correlation with subsequent infection (p < 0.05).
In three (7%) of forty-two patients with Gustilo grade-I, II, and IIIA fractures, wound infections developed and were successfully treated with antibiotics. Five patients with delayed union required a revision open reduction and internal fixation with bone-grafting.
Seventeen patients had grade-IIIB fractures. In two of these patients, wound infections developed. With the numbers studied, the infection rate of 12% did not differ significantly from that of patients with grade-I, II and IIIA fractures (p = 0.571). One patient required a below-the-knee amputation. With the numbers studied, patients requiring flap coverage did not demonstrate a significantly different ankle-specific outcome or generalized outcome compared with patients with grade-I, II, and IIIA fractures.
Published data pertaining to the treatment of open pilon fractures are found predominantly in small series of open injuries, or as a subset of series including both open and closed fractures7,16,21. In the present series, we assessed the outcome for all patients with an open pilon fracture treated acutely with open reduction and internal fixation at a single facility over an eleven-year period. To our knowledge, this is the largest reported series of open pilon fractures so treated by a single surgeon.
The eighty-four pilon fractures in the series described by Rüedi and Allgöwer included only three open fractures5. They reported a combined 12% rate of superficial wound infection and a 5% rate of deep infection. Subsequent authors, including Helfet et al.30 and Sirkin et al.3, reported an acceptable rate of soft-tissue complications if definitive fixation of the fracture is appropriately delayed. In a series of fifty-six pilon fractures that included twenty-two open injuries, Sirkin et al. reported an overall rate of infection of 5.3% and an 11% rate in the subset of patients with an open fracture. Conroy et al. reported on a series of thirty-four grade-IIIB open pilon fractures that were managed with early open reduction and internal fixation and free tissue transfer; an infection developed in 18.7% of the patients and amputation was required in 6.2%31. Chen et al. reviewed the outcome of a series of thirty-eight open and ninety closed fractures treated with delayed open reduction and internal fixation32. Those authors reported a 5.5% rate of deep infection and a 3.1% rate of osteomyelitis.
In the current series, two (3%) of fifty-nine patients were diagnosed as having a deep wound infection. No patient had development of chronic osteomyelitis. One patient had positive cultures at the time of a planned bone graft, and another had an infection develop following hardware removal at two years after the injury. Both patients were successfully treated with a six-week course of culture-specific intravenous antibiotics. Three patients (5%) were treated with oral antibiotics for a presumed superficial infection. We attribute this lower infection rate to aggressive initial débridement and provisional stabilization, routine second-look débridement, subsequent débridement as needed, stable definitive fixation, and an appropriate antibiotic regimen. We also maintained a low threshold for free tissue transfer as we believe that vascularized free flaps obliterate dead space and increase local blood flow, thus aiding wound-healing and decreasing the risk of infection33,34.
Alternative treatment modalities based on the definitive use of external fixation alone or in combination with limited internal fixation have theoretical advantages. External fixation may be accomplished rapidly, causing limited additional soft-tissue injury; however, the proximity of the injury to the joint often mandates that the joint be spanned. This increases the likelihood of tibiotalar or subtalar joint stiffness, or both18,35-38. In addition, prolonged use of external fixation is associated with pin-track infection, and the proximity of Schanz pins or fine wires to the joint increases the likelihood of septic arthritis39,40. When supplemental internal fixation is required, the risks associated with an open procedure are also present. Hybrid and ring fixators incorporating periarticular fine wires have been used in an attempt to reduce the articular component of the fracture, avoid immobilization of the tibiotalar and subtalar joints, and allow mobilization. Limitations inherent in this technique include difficulty in achieving and/or confirming the articular reduction, difficulty in visualizing the articular surface with the fixator in place, and an inability to dynamize the construct4,41-45.
The optimal long-term outcome following an open pilon fracture requires that soft-tissue complications be avoided and osseous anatomy be restored. The restoration of limb alignment and rotation, along with anatomic restoration of the joint surface, is imperative as the development of posttraumatic arthritis has been shown to correlate closely to the severity of injury and the quality of reduction12. The technique described here allows for direct visualization and optimal reduction of both the articular surface and the metadiaphyseal component of the injury. Our review suggests that open reduction and internal fixation can be performed and soft-tissue complications minimized through the use of the described protocol. With this protocol, an adequate fracture reduction was achieved and maintained in all patients. Radiographic follow-up did not include an assessment of tibiotalar or subtalar arthritis; however, the clinical outcome, including the modified Mazur score and the physical component score of the SF-36v2, were suggestive of arthritis in a high proportion of patients. Despite the patients’ physical limitations and pain, the scores on the mental component of the SF-36v2 were similar to those of age-matched controls in the majority of the age groups. We attribute this finding to the patients being appropriately educated as to the severity of the injury and having realistic expectations.
This retrospective study was conducted in a select cohort of patients with high-energy open pilon fractures. Patient demographic and fracture pattern characteristics were random and not specified by the protocol. A prospective, randomized, controlled trial comparing different methods of definitive fixation with appropriate outcome assessment tools and radiographic assessment of ankle and subtalar joint arthritis is necessary to better define the components of an optimal treatment protocol.
Despite the above noted limitations, we believe that open reduction and internal fixation can be performed safely in high-energy open pilon fractures when a staged protocol that emphasizes respect for soft-tissue injury is used.