A thirty-nine-year-old man (290 lb [131.5 kg] and 6 ft, 2 in [1.9 m]) was the restrained driver of a truck involved in a high-speed collision with another motor vehicle. He sustained intraabdominal injuries and multiple musculoskeletal injuries, including a left radial styloid fracture, a left supracondylar femoral fracture, a fracture of the medial process of the right talus, and a left total talar extrusion (without a fracture) through a large lateral ankle wound (Fig. 1). The talus was transported to the hospital by the ambulance service in a sterile plastic bag on ice. Once in the hospital, the talus was kept on ice in a saline-solution bath containing cefazolin. The neurovascular status of the extremity was intact.
The patient was taken to the operating room, where the wound edges were débrided of devitalized and contaminated tissue and the open wound was liberally irrigated with 9 L of normal saline solution. The extruded talus was examined, and slight impaction of 50% of the medial talar head articular surface was noted. There were no fractures of the talar neck, body, or processes. After débridement and washing with saline solution and cefazolin, the talus was reimplanted through the open wound and reduced into position. Congruency of the tibiotalar, talonavicular, and subtalar joints was confirmed on fluoroscopy. Provisional fixation was achieved with a single Kirschner wire. No attempt was made to repair the deep deltoid. The wound was packed with antibiotic beads (Simplex bone cement [Stryker, Hamilton, Ontario, Canada] with tobramycin, to which vancomycin had been added) and left open. The time from the injury to the reimplantation of the talus was four hours. A repeat irrigation and débridement was performed at seventy-two hours. The provisional Kirschner wire was removed at that time, and the wound was closed. The antibiotic therapy was continued for seven days, and the patient was discharged from the hospital ten days postinjury. No intraoperative or early postoperative complications occurred.
The patient remained non-weight-bearing on the left lower extremity for ten weeks; this was followed by progressive weight-bearing in a walker boot. Full weight-bearing was achieved at fifteen weeks. He participated in physiotherapy for six months, which was limited by ongoing knee pain attributed to the retrograde nail fixation of the supracondylar femoral fracture. He was followed clinically, radiographically, and with serial magnetic resonance imaging scans at two weeks, six months, and two and five years. The magnetic resonance imaging protocol consisted of standard axial, sagittal, and coronal proton-density images with additional sagittal and coronal postgadolinium fat-saturated T2-weighted sequences at five years.
At five years after the injury, the patient could walk for two hours without aids and had a mild limp and moderate hindfoot pain. He could stand for four hours. He did not participate in sports before or after the injury, and he was able to return to sedentary employment. There was no specific deterioration in his clinical course over the five years of follow-up. The overall alignment of the left ankle, hindfoot, and midfoot was normal. The range of motion of the ankle was 25° of plantar flexion and 10° of dorsiflexion. Subtalar motion was very restricted.
Radiographs of the left ankle made immediately after the surgery and at two and five years are shown in Figure 2. Progressive radiographic evidence of arthritic changes at the talonavicular joint as well as about the posterior facet of the subtalar joint and the tibiotalar joint was seen. There was also increased sclerosis of the talar body over time, suggesting avascularity. No collapse of the body was evident.
Serial magnetic resonance images acquired at two weeks, six months, two years, and five years after the injury are shown in Figure 3. At two weeks, there was no magnetic resonance imaging evidence of substantial signal change in the talus relative to the adjacent tibia and calcaneus, indicating an absence of acute inflammatory response within the talus. At six months postinjury, there was mild signal abnormality throughout the talus, indicating some physiologic activity, despite the talus having been stripped of all of its blood supply at the time of injury. There was no magnetic resonance imaging evidence of osteonecrosis of the talus at either two weeks or six months after the injury. At two years postinjury, the talus showed diffuse signal changes, consistent with early osteonecrosis, but no collapse was evident.
At five years postinjury (Figs. 3-D and 3-E), the diffuse signal changes seen at two years had become localized to the central portion of the talar body. The anterior talar signal characteristics were identical to those of the adjacent tarsal bones, suggesting that the head and neck had been progressively revascularized. There was a clear demarcation between the normal anterior aspect of the talus and the central portion, with enhancement of the sclerotic border. This pattern of enhancement was highly suggestive of osteonecrosis. The talar dome continued to have a normal contour, and no collapse was identified.
The most extreme variant of talar dislocation is an extruded talus. Guidance for the treatment of an extruded talus is limited to that derived from small case series, which is not surprising given the rarity of this injury1-12. When treating a patient with an extruded talus, the orthopaedic surgeon must weigh the benefit of maintaining the anatomic integrity of the tibiotalar and subtalar joints through reimplantation of the talus against the acute and long-term risks of infection, osteonecrosis, and posttraumatic arthritis that may be associated with this procedure1-12.
Primary tibiocalcaneal fusion has been recommended by some authors7-10. However, this procedure inevitably results in a leg-length discrepancy, and complete loss of ankle and hindfoot motion, or a pseudarthrosis, all of which impose substantial functional limitations and limit future reconstructive options. Therefore, most authors recommend reimplantation of the native talus, with arthrodesis and talectomy reserved for treatment of post-reimplantation complications1-6.
We are not aware of any previous study in which the progression of an acutely reimplanted talus was followed with serial magnetic resonance imaging scans. Since magnetic resonance imaging is the most sensitive test available to determine the presence and extent of osteonecrosis and to look for evidence of revascularization and collapse, serial scans may provide useful information13.
The risk of osteonecrosis and the rate at which it progresses following a total talar dislocation without a fracture remain largely unknown. Literature describing Hawkins type-IV talar neck fractures suggests a risk of osteonecrosis approaching 100%, given the damage to all of the vascular structures that occurs with a total talar extrusion14,15. However, some authors have found that not all extruded tali without a fracture go on to collapse following reimplantation, leading them to suggest that osteonecrosis had not developed in these cases2,3,5,6,12. These authors did not provide magnetic resonance imaging evidence to support this suggestion. It is clear that the talus in our patient was completely stripped of all of its blood supply, rendering it completely avascular at the time of injury. It would therefore seem intuitive that osteonecrosis was inevitable; however, osteonecrosis did not become evident on magnetic resonance imaging until two years postinjury. At five years postinjury, the talar head and neck appeared to have been revascularized and the area of osteonecrosis had become localized to the central body. Despite the avascularity in the body, the talus had not collapsed. This finding suggests that the rate of progression and the extent of clinically evident osteonecrosis following talar extrusions with a talar neck fracture differ from those following talar extrusions without a fracture.
The serial magnetic resonance imaging data presented here provide new information that supports this difference in progression rates of osteonecrosis. The typical magnetic resonance imaging changes indicating osteonecrosis after talar neck fracture have been shown to occur within ten weeks after the injury, and subsequent collapse or revascularization occurs within another thirty-six weeks16. In contrast, in our case, magnetic resonance imaging changes consistent with osteonecrosis were not seen until two years postinjury, with slow progression for up to five years after the injury, and radiographic changes were not evident until well after two years. Therefore, our magnetic resonance imaging data strongly suggest that revascularization occurs much more slowly when the talar extrusion is not associated with a talar neck fracture.
The pathophysiology underlying the observed finding of slower revascularization and progression of the sequelae of osteonecrosis following reimplantation of an extruded talus that had not fractured remains unknown. However, a possible explanation is that the large surface area at the site of a talar neck fracture offers a much greater opportunity for entry of new vascular channels into the talar body compared with the small area for potential vascular reentry on the smooth cortical bone of an intact talar neck. Also, the remaining soft-tissue attachments at the site of a fracture-dislocation may provide a blood supply that is adequate to locally distribute hematogenous inflammatory mediators, hastening bone resorption and subsequent collapse. The lack of any soft-tissue attachment following a complete extrusion may result in a suppressed local inflammatory response and slower collapse.
Our patient was able to return to his preinjury level of activity, and he described minimal disability related to the ankle despite magnetic resonance imaging changes consistent with osteonecrosis without collapse. These findings indicate that satisfactory medium-term, and potentially long-term, outcomes are possible after such a severe injury and that the need for late reconstructive surgery is not absolute. This observation supports the finding of previous authors that osteonecrosis of the talus does not necessarily lead to an unsatisfactory result3,12,13 and also supports a recommendation of reimplantation of a totally extruded talus when all conditions, such as a lack of contamination or articular damage, are favorable.