FIGURE 32-1 Salter–Harris anatomic classification as applied to injuries of the distal tibial epiphysis.

Injury classifications based upon the mechanism of injury may have some advantages. The description of the injury includes the anatomic deformity and the forces that produced the injury. An understanding of these forces can facilitate reduction of a displaced fracture. Advanced imaging techniques that allow for comprehensive three-dimensional visualization of the fracture anatomy also facilitate surgical planning and reduction techniques.

Both anatomical and mechanism-of-injury classifications can provide useful information for determining appropriate treatment. The prognoses for growth and deformity have been predicted on the basis of both types of classification.^98,99,196 A theoretical advantage of mechanism-of-injury classifications is that identification of the force producing the injury might give even more information about the possible development of growth arrest than anatomical classifications. For example, a Salter–Harris type III or IV fracture of the tibia produced by a shearing or crushing force might be more likely to result in growth arrest than is a similar injury produced by an avulsion force (Fig. 32-2). However, it is difficult to establish that one type of classification is superior to the other in this regard because of the relatively small numbers of patients reported, the varying ages of patients in most series, and questions about the reproducibility of various classifications.

Ideally, classification systems should have high interobserver and intraobserver agreement. Thomsen et al.²⁰⁴ studied the reproducibility of the Lauge-Hansen (mechanism-of-injury) and Weber (anatomical) classifications in a series of adult ankle fractures. After all investigators in the study had received a tutorial on both systems and their application, they were asked to classify 94 fractures. On the first attempt, only the Weber classification produced an acceptable level of interobserver agreement. On a second attempt, the Weber classification and most of the Lauge-Hansen classification achieved an acceptable level of interobserver agreement. These authors concluded that all fracture classification systems should have demonstrably acceptable interobserver agreement rates before they are adopted, an argument made even more forcefully in an editorial by Burstein.³³ Vahvanen and Aalto²⁰⁷ compared their ability to classify 310 ankle fractures in children with the Weber, Lauge-Hansen, and Salter–Harris classifications. They found that they were “largely unsuccessful” using the Weber and Lauge-Hansen classifications, but could easily classify the fractures using the Salter–Harris system.

The most widely accepted mechanism-of-injury classification of ankle fractures in children is that described by Dias and Tachdjian,⁵⁷ who modified the Lauge-Hansen classification based on their review of 71 fractures (Fig. 32-3). Their original classification (1978) consisted of four types in which the first word refers to the position of the foot at the time of injury and the second word refers to the force that produces the injury.

Other fracture types were subsequently added, including axial compression, juvenile Tillaux, triplane, and other physeal injuries by Tachdjian.²⁰¹ Syndesmosis injuries have also been recently described.⁵¹ “Axial compression injury” describes the mechanism of injury but not the position of the foot. Juvenile Tillaux and triplane fractures are called transitional fractures as they occur when the physis is transitioning from open to closed, and are believed to be caused by external rotation.¹⁷⁵ The final category, “other physeal injuries,” includes diverse injuries, many of which have no specific mechanism of injury.

Classification of Ankle Fracture in Children (Dias–Tachdjian) (Fig. 32-3)

Supination–Inversion

• Grade I: The adduction or inversion force avulses the distal fibular epiphysis (Salter–Harris type I or II fracture). Occasionally, the fracture is transepiphyseal; rarely, the lateral ligaments fail.

• Grade II (Fig. 32-4): Further inversion produces a tibial fracture, usually a Salter–Harris type III or IV and rarely a Salter–Harris type I or II injury, or the fracture passes through the medial malleolus below the physis (Fig. 32-5).

Supination–Plantarflexion

The plantarflexion force displaces the epiphysis directly posteriorly, resulting in a Salter–Harris type I or II fracture. Fibular fractures were not reported with this mechanism. The tibial fracture may be difficult to see on anteroposterior radiographs (Fig. 32-6).

Supination–External Rotation

• Grade I: The external rotation force results in a Salter–Harris type II fracture of the distal tibia (Fig. 32-7). The distal fragment is displaced posteriorly, as in a supination–plantarflexion injury, but the Thurston–Holland fragment is visible on the anteroposterior radiographs, with the fracture line extending proximally and medially. Occasionally, the distal tibial epiphysis is rotated but not displaced.

• Grade II: With further external rotation, a spiral fracture of the fibula is produced, running from anteroinferior to posterosuperior (Fig. 32.8).

Pronation–Eversion–External Rotation

A Salter–Harris type I or II fracture of the distal tibia occurs simultaneously with a transverse fibular fracture. The distal tibial fragment is displaced laterally and the Thurston–Holland fragment, when present, is lateral or posterolateral (Fig. 32-9). Less frequently, a transepiphyseal fracture occurs through the medial malleolus (Salter–Harris type II). Such injuries may be associated with diastasis of the ankle joint, which is uncommon in children.

Axial Compression

This results in a Salter–Harris type V injury of the distal tibial physis. Initial radiographs usually show no abnormality, and the diagnosis is established when growth arrest is demonstrated on follow-up radiographs.

Transitional Fractures of the Distal Tibia and Fibula

Because the distal tibial physis closes in an asymmetric pattern over a period of about 18 months, injuries sustained during this period can produce fracture patterns that are not seen in younger children with completely open physes.¹²⁶ This group of fractures has been labeled “transitional” fractures because they occur during the transition from a skeletally immature ankle to a skeletally mature ankle. Such fractures, which include juvenile Tillaux and “triplane” fractures with two to four fracture fragments, have been described by Kleiger and Mankin,¹⁰⁸ Marmor,¹³¹ Cooperman et al.,⁴⁶ Kärrholm et al.,⁹⁷ and Denton and Fischer.⁵⁴ The adolescent pilon fracture has been described by Letts et al.¹¹⁷ The incisural fracture has been described by Cummings and Hahn.⁵² Syndesmosis injuries have been described by Cummings.⁵¹

Classification of these fractures is even more confusing than that of other distal tibial fractures. Advocates of mechanism-of-injury systems agree that most juvenile Tillaux and triplane fractures are caused by external rotation, but they disagree as to the position of the foot at the time of the injury.^55,56,164 Some authors⁵⁶ classify juvenile Tillaux fractures as stage I injuries, with further external rotation causing triplane fractures, and still further external rotation causing stage II injuries with fibular fracture. Others emphasize the extent of physeal closure as the only determinant of fracture pattern.⁴⁵

Advocates of anatomical classifications are handicapped by the different anatomical configurations triplane fractures may exhibit on different radiograph projections, making tomography, computed tomography (CT) scanning, or examination at open reduction necessary to determine fracture anatomy and number of fragments. Because these fractures occur near the end of growth, growth disturbance is rare. Anatomical classification is, therefore, more useful for descriptive purposes than for prognosis.

Juvenile Tillaux Fracture of the Distal Tibia and Fibula

The juvenile Tillaux fracture is a Salter–Harris type III fracture involving the anterolateral distal tibia. The portion of the physis not involved in the fracture is closed (Fig. 32-10).

Triplane Fracture of the Distal Tibia and Fibula

A group of fractures that have in common the appearance of a Salter–Harris type III fracture on the anteroposterior radiographs and of a Salter–Harris type II fracture on the lateral radiographs (Fig. 32-11). CT scans can be very helpful to understand the complex anatomy of these fractures (Fig. 32-11).^14,49,107 Ipsilateral triplane and diaphyseal fractures have been reported, and one of the fractures can be missed if adequate images are not obtained.^14,49,91

Adolescent Pilon Fractures of the Distal Tibia and Fibula

The pediatric/adolescent pilon fracture¹¹⁷ is defined as a fracture of the “tibial plafond with articular and physeal involvement, variable talar and fibular involvement, variable comminution, and greater than 5 mm of displacement” (Fig. 32-12). Based upon a small number of cases, Letts et al. developed a three-part classification system. Type I fractures have minimal comminution and no physeal displacement. Type II fractures have marked comminution and less than 5 mm of physeal displacement. Type III fractures have marked comminution and more than 5 mm of physeal displacement.

Incisura Fractures of the Distal Tibia and Fibula

Incisural fractures are fractures that resemble Tillaux on standard radiographs, but the size of the fragment is smaller than that typically seen with the Tillaux fractures (Fig. 32-13).⁵² On the CT scan, this fracture does not extend to the anterior cortex of the distal tibia (Fig. 32-14). The mechanism of injury may be an avulsion of the fragment by the interosseous ligament. This may be a variant of an adult tibiofibular diastasis injury.

Syndesmosis Injuries of the Distal Tibial and Fibular Fractures

The authors have seen syndesmosis injuries in young patients. These have been associated with fractures of the distal fibula, Tillaux injuries, S-H I fractures, and proximal fibula fractures (Figs. 32-15, 32-16, 32-17). These fractures are probably rare and there is very limited literature on this injury.¹⁵⁶

Stress Fractures of the Distal Tibia and Fibula

Stress fractures can occur in the distal tibial metaphyseal area (Fig. 32-18), or through the distal fibular physis (Fig. 32-19). These patients may present with warmth, swelling, and pain around the metaphyseal or physeal regions. In our experience, these injuries are more common in gymnasts, ice skaters, and running/endurance athletes. We have seen stress fractures through the distal fibular physeal scar in running athletes.

Signs and Symptoms of Distal Tibial and Fibular Fractures

Patients with significantly displaced fractures have severe pain and obvious deformity. The position of the foot relative to the leg may provide important information about the mechanism of injury (Fig. 32-20) and should be considered in planning reduction. The status of the skin, pulses, and sensory and motor function should be determined and recorded. Tenderness, swelling, and deformity in the ipsilateral leg and foot should be noted. In patients with tibial shaft fractures, the ankle should be carefully evaluated clinically and radiographically.

Although compartment syndromes are rare, they do occur in these locations.^47,139 If patients are admitted to the hospital, discussion with the nursing staff about signs and symptoms of compartment syndrome is important. If patients are treated as outpatients, the patient and family should be informed about the possibility of compartment syndrome and instructed to return to the hospital for evaluation if problems with pain control develop.

Imaging and Other Diagnostic Studies for Distal Tibial and Fibular Fractures

Patients with nondisplaced or minimally displaced ankle fractures often have no deformity, minimal swelling, and moderate pain. Because of their benign clinical appearance, such fractures may be easily missed if radiographs are not obtained. Petit et al.¹⁶¹ reviewed 2,470 radiographs from pediatric emergency rooms, demonstrating abnormal radiographic findings in 9%. Guidelines known as The Ottawa Ankle Rules have been established for adults to try to determine which injuries require radiographs.¹⁹⁹ The Ottawa Ankle Rules have also been evaluated in children over the age of 5. These rules appear to be a reliable tool to exclude fractures in children greater than 5 years of age presenting with ankle and midfoot injuries and may significantly decrease x-ray use with a low likelihood of missing a fracture.⁵⁹ The indications for radiographs according to the guidelines are complaints of pain near a malleolus with either inability to bear weight or tenderness to palpation at the malleolus. Chande⁴³ prospectively studied 71 children with acute ankle injuries to determine if these guidelines could be applied to pediatric patients with ankle injuries. It was determined that if radiographs were obtained only in children with tenderness over the malleoli, and inability to bear weight, a 25% reduction in radiographic examinations could be achieved without missing any fractures. The physical examination should focus upon physeal areas of the tibia and fibula, when evaluating ankle injuries, to determine if radiographs are necessary. Interpretation of radiographs should focus upon signs of physeal injury, including soft tissue swelling in these regions.

For patients with obvious deformities, anteroposterior, mortise, and lateral radiographs centered over the ankle may provide sufficient information to plan treatment. Although obtaining views of the joint above and below is recommended for most fractures, obtaining a film centered over the midtibia to include the knee and ankle joints on the radiographs significantly decreases the quality of ankle views and is not recommended.

For patients without obvious deformities, a high-quality mortise view of the ankle is essential in addition to anteroposterior and lateral views. On a standard anteroposterior view, the lateral portion of the distal tibial physis is usually partially obscured by the distal fibula. The vertical component of a triplane or Tillaux fracture can be hidden behind the overlying fibular cortical shadow.¹¹⁹ A study by Vangsness et al.²⁰⁸ found that diagnostic accuracy was essentially equal when using anteroposterior, lateral, and mortise views compared with using only mortise and lateral views. Therefore, if only two views are to be obtained, the anteroposterior view may be omitted and lateral and mortise views obtained.

Haraguchi et al.,⁷⁹ described two special views designed to detect avulsion fractures from the lateral malleolus that are not visible on routine views, and to distinguish whether they represent avulsions of the anterior tibiofibular ligament or the calcaneofibular ligament attachments. The anterior tibiofibular ligament view is made by positioning the foot in 45 degrees of plantarflexion and elevating the medial border of the foot 15 degrees. The calcaneofibular ligament view is obtained by rotating the leg 45 degrees inward.

Stress views are occasionally recommended historically to rule out ligamentous instability, but the authors see only rare indications for stress radiography in skeletally immature patients. The discomfort of stress views in an acute injury can be avoided by using other imaging options, such as magnetic resonance imaging (MRI).

Bozic et al.²⁷ studied the age at which the radiographic appearance of the incisura fibularis, tibiofibular clear space, and tibiofibular overlap develops in children.²⁷ The purpose of their study was to facilitate the diagnosis of distal tibiofibular syndesmotic injury in children. They found that the incisura became detectable at a mean age of 8.2 years for girls and 11.2 years for boys. The mean age at which tibiofibular overlap appeared on the AP view was 5 years for both sexes; on the mortise view, it was 10 years for girls and 16 years for boys. The range of clear space measurements in normal children was 2 to 8 mm, with 23% of children having a clear space greater than 6 mm—a distance considered abnormal in adults.

CT is useful in the evaluation of intra-articular fractures, especially juvenile Tillaux and triplane fractures (Fig. 32-21).^{7,14,31,49,66,95,107} Transverse images are obtained with thin cuts localized to the joint, and high-quality reconstructions can be produced in the coronal and sagittal planes without repositioning the ankle. Three-dimensional CT reconstructions may add further useful information, and readily available software packages allow easy production of such images (Fig. 32-22). These images can assist with minimally invasive approaches, the use of percutaneous reduction clamps, and positioning of fixation screws.

MRI may be useful in the evaluation of complex fractures of the distal tibia and ankle in patients with open physes. Smith et al.¹⁹⁵ found that of four patients with acute (3 to 10 days) physeal injuries, MRI showed that three had more severe fractures than indicated on plain films (Fig. 32-23). Early MRI studies (3 to 17 weeks after injury) not only added information about the pattern of physeal disruption but also supplied early information about the possibility of growth abnormality. MRI has been reported to be occasionally helpful in the identification of osteochondral injuries to the joint surfaces in children with ankle fractures.¹⁰⁵ Although these injuries may be more common in adult fractures, we believe that these types of injuries are very rare in younger patients.

Carey et al.³⁶ obtained MRI studies on 14 patients with known or suspected growth plate injury. The MRI detected five radiographically occult fractures in the 14 patients, changed the Salter–Harris classification in two cases, and resulted in a change in treatment plan in 5 of the 14 patients studied. These studies would seem to contradict an earlier study by Petit et al.,¹ that showed only one patient in a series of 29 patients in whom MRI revealed a diagnosis different from that made on plain films. Iwinska-Zelder et al.⁹⁰ found that the MRI changed the management in 4 of 10 patients with ankle fractures seen on plain radiographs. Seifert et al.¹⁸⁷ found the MRI identified physeal injuries that were not identified by plain radiographs. At this time, the indications for MRI in the evaluation of ankle fractures in skeletally immature patients are still being defined, but this imaging modality may be a more sensitivity tool for identification of minimally displaced or more complex injuries.¹⁴ In a recent prospective study of skeletally immature patients with clinically diagnosed Salter I fractures of the distal fibula, none of the 18 patients imaged by MRI had evidence of physeal injury. The patients had a mean age of 8 years, and over 70% had evidence of ligamentous sprain on MRI. This questions the principle that the physis is the weak link in the musculoskeletal system in this age group.²⁴ If physeal arrest occurs, MRI scans are useful for mapping physeal bars.^69,82

The use of ultrasound to detect radiographically occult fractures may be used for pediatric ankle fractures.¹⁹²

Pitfalls in Diagnosis

A number of accessory ossification centers and normal anatomical variations may cause confusion in the interpretation of plain films of the ankle (Fig. 32-24). In a group of 100 children between the ages of 6 and 12 years, Powell¹⁶⁵ found accessory ossification centers on the medial side (os subtibiale) in 20% and on the lateral side (os subfibulare) in 1%. If they are asymptomatic on clinical examination, these ossification centers are of little concern, but tenderness localized to them may indicate an injury. Stress views to determine motion of the fragments or MRI scanning may occasionally be considered if an injury to an accessory ossification center is suspected.

Clefts in the lateral side of the tibial epiphysis may simulate juvenile Tillaux fractures, and clefts in the medial side may simulate Salter–Harris type III fractures.¹⁰⁴ The presence of these clefts on radiographs of a child with an ankle injury may result in overtreatment if they are misdiagnosed as a fracture. Conversely, attributing a painful irregularity in these areas to anatomical variation may lead to undertreatment (Fig. 32-25). Other anatomical variations include a bump on the distal fibula that simulates a torus fracture and an apparent offset of the distal fibular epiphysis that simulates a fracture. These radiographic findings should be correlated with physical examination findings of focal swelling and point tenderness that correspond with the imaging in the diagnosis of skeletal injury.

PATHOANATOMY AND APPLIED ANATOMY RELATING TO DISTAL TIBIAL AND FIBULAR FRACTURES

The ankle joint closely approximates a hinge joint. It is the articulation between the talus and the ankle mortise, which is a syndesmosis consisting of the distal tibial articular surface, the medial malleolus, and the distal fibula or lateral malleolus.

Four ligamentous structures bind the distal tibia and fibula into the ankle mortise (Fig. 32-26). The anterior and posterior-inferior tibiofibular ligaments course inferiorly from the anterior and posterior surfaces of the distal lateral tibia to the anterior and posterior surfaces of the lateral malleolus. The anterior ligament is important in the pathomechanics of “transitional” ankle fractures. Just anterior to the posterior-inferior tibiofibular ligament is the broad, thick inferior transverse ligament, which extends down from the lateral malleolus along the posterior border of the articular surface of the tibia, almost to the medial malleolus. This ligament serves as a part of the articular surface for the talus. Between the anterior and posterior-inferior tibiofibular ligaments, the tibia and fibula are bound by the interosseous ligament, which is continuous with the interosseous membrane above. This ligament may be important in the pathomechanics of what we have termed incisural fractures.

On the medial side of the ankle, the talus is bound to the ankle mortise by the deltoid ligament (Fig. 32-27). This ligament arises from the medial malleolus and divides into superficial and deep layers. Three parts of the superficial layer are identified by their attachments: Tibionavicular, calcaneotibial, and posterior talotibial ligaments. The deep layer is known as the anterior talotibial ligament, again reflecting its insertion and origin. On the lateral side, the anterior and posterior talofibular ligament, with the calcaneofibular ligaments, make up the lateral collateral ligament (Fig. 32-28).

In children, all medial and lateral ligaments originate distal to the tibial or fibular physis. Because the ligaments are often stronger than the physes, physeal fractures have generally been viewed as more common than ligamentous injuries in children. Advanced imaging studies have shown that the rate of ankle fractures compared to ligamentous injuries is variable,⁶³ and this is likely dependent on multiple factors such as the mechanism of injury, rate of force application, relative strength of the physis, and age of the patient. When distal tibia and fibular fragments are displaced together, the syndesmosis at the level of the fracture is usually intact (Fig. 32-29).

The distal tibial ossification center generally appears at 6 to 24 months of age. Its malleolar extension begins to form around the age of 7 or 8 years and is mature or complete at the age of 10 years. The medial malleolus develops as an elongation of the distal tibia ossific nucleus, although in 20% of cases, this may originate from a separate ossification center, the os tibial. This can be mistaken as a fracture.¹⁰² The physis usually closes around the age of 15 years in girls and 17 years in boys. This process takes approximately 18 months and occurs first in the central part of the physis, extending next to the medial side, and finally ending laterally. This asymmetric closure sequence is an important anatomical feature of the growing ankle and is responsible for certain fracture patterns in adolescents, especially transitional fractures (Fig. 32-30).

The distal fibular ossification center appears around the age of 9 to 24 months. This physis is located at the level of the ankle joint initially, and moves distally with growth.^100,214 Closure of this physis generally follows closure of the distal tibial physis by 12 to 24 months.

The locations of the sensory nerves are important anatomic landmarks, as surgical exposures should aim to protect these structures. The superficial peroneal nerve branches may be most vulnerable around the ankle, especially during arthroscopic and arthrotomy approaches for triplane and Tillaux fractures.¹⁵ This is important when arthroscopic and percutaneous reduction techniques are employed for fracture treatment (refer to section on Fracture Reduction Tips, Arthroscopic Assistance, Use of Percutaneous Clamps, Implants).

TREATMENT OPTIONS FOR DISTAL TIBIAL AND FIBULAR FRACTURES

Appropriate treatment of ankle fractures in children depends on the location of the fracture, the degree of displacement, and the age of the child (Table 32-1). Nondisplaced fractures may be simply immobilized. A recent randomized clinical trial for minimally displaced low-risk ankle fractures compared a fiberglass posterior splint to a removable ankle stirrup brace. This study demonstrated good outcomes in both groups.⁸ Closed reduction and cast immobilization may be appropriate for displaced fractures; if the closed reduction cannot be maintained with casting, skeletal fixation may be necessary. If closed reduction is not possible, open reduction may be indicated, followed by internal fixation or cast immobilization.

The anatomical type of the fracture (usually defined by the Salter–Harris classification), the mechanism of injury, and the amount of displacement of the fragments are important considerations. When the articular surface is disrupted, the amount of articular step-off or separation must be measured. The neurologic and vascular status of the limb or the status of the skin may require emergency treatment of the fracture and associated problems. The general health of the patient and the time since injury must also be considered.

Distal Tibial Fractures

Salter–Harris Type I and II Fractures

According to Dias and Tachdjian,^57,201 Salter–Harris type I fractures of the distal tibia can be caused by any of the four mechanisms: Supination–inversion, supination–plantarflexion, supination–external rotation, or pronation–eversion–external rotation. Spiegel et al.¹⁹⁶ reported that these fractures accounted for 15.2% of 237 ankle injuries in their series and occurred in children significantly younger (average age, 10.5 years) than those with other Salter–Harris types of fractures.

The mechanism of injury is deduced primarily by the direction of displacement of the distal tibial epiphysis; for example, straight posterior displacement indicates a supination–plantarflexion mechanism. The type of associated fibular fracture is also indicative of the mechanism of injury; for example, a high, oblique or transverse fibular fracture indicates a pronation–eversion–external injury, whereas a lower spiral fibular fracture indicates a supination–external rotation injury. Lovell,¹²⁴ Broock and Greer,²⁷ and Nevelös and Colton¹⁴⁵ reported unusual Salter–Harris type I fractures in which the distal tibial epiphysis was externally rotated 90 degrees without fracture of the fibula or displacement of the tibial epiphysis in any direction in the transverse plane.

Cast immobilization is generally sufficient treatment for nondisplaced Salter–Harris type I fractures of the distal tibia. A below-knee cast worn for 3 to 4 weeks may suffice, with the first 2 to 3 weeks limited to nonweight bearing. An above-knee cast may also be used, although this may not be necessary as these fractures are usually very stable. In very active patients who may not comply with activity/weight-bearing restrictions, this type of cast may be an advantage. After cast removal, use of a removable leg/ankle walking boot may be used, followed by a therapy program in older patients or those trying to return to competitive sports at an earlier time. In our experience, formal supervised therapy is not necessary in younger patients. The normal activity of these children is usually sufficient therapy.

Most displaced fractures can be treated with closed reduction and cast immobilization. An above-knee non–weight-bearing cast is preferable initially, as this should reduce the risk of displacement after reduction. These casts may be changed to a short-leg walking cast or removable walking boot at 3 to 4 weeks. These fractures can displace in the first 1 to 2 weeks postoperatively, and close follow-up with radiographic surveillance for this is necessary. One of the authors (KS) frequently places one or two Kirschner wires at the time of closed reduction, to prevent displacement after reduction under anesthesia (Fig. 32-31). These pins are usually removed in the clinic 2 to 3 weeks after placement. Under these circumstances, a below-knee cast can be used.

Salter–Harris Type II Fractures

Salter–Harris type II fractures can also be caused by any of the four mechanisms of injury described by Dias and Giegerich.⁵⁶ In the series of Spiegel et al.,¹⁹⁶ Salter–Harris type II fractures were the most common injuries (44.8%). In addition to the direction of displacement of the distal tibial epiphysis and the nature of any associated fibular fracture, the location of the Thurston–Holland fragment is helpful in determining the mechanism of injury; for example, a lateral fragment indicates a pronation–eversion–external rotation injury; a posteromedial fragment, a supination–external rotation injury; and a posterior fragment, a supination–plantarflexion injury (Fig. 32-32).

Nondisplaced fractures can be treated with cast immobilization usually with an above-knee cast for 3 to 4 weeks, followed by a below-knee walking cast or removable cast/walking boot for another 3 to 4 weeks.

Although most authors agree that closed reduction of significantly displaced Salter–Harris type II ankle fracture should be attempted, opinions differ as to what degree of residual displacement or angulation is unacceptable and requires open reduction. Based on follow-up of 33 Salter–Harris type II ankle fractures, Carothers and Crenshaw³⁷ concluded that “accurate reposition of the displaced epiphysis at the expense of forced or repeated manipulation or operative intervention is not indicated since spontaneous realignment of the ankle occurs even late in the growing period.” They found no residual angulation at follow-up in patients who had up to 12 degrees of tilt after reduction, even in patients as old as 13 years of age at the time of injury. Spiegel et al.,¹⁹⁶ however, reported complications at follow-up in 11 of 16 patients with Salter–Harris type II ankle fractures. Because 6 of these 11 patients had angular deformities that were attributed to lack of adequate reduction of the fracture, Spiegel et al. recommend “precise anatomical reduction.”

Barmada et al.⁷ reviewed a series of Salter–Harris type I and II fractures. In patients with more than 3 mm of physeal widening, the risk of premature physeal closure was 60%, compared with 17% in patients with less than 3 mm of physeal widening. Although they were unable to demonstrate a significant decrease in partial physeal arrest in those treated with surgery, they recommended open reduction and removal of the entrapped periosteal flap. Leary et al.¹¹³ studied 15 distal tibia fractures with premature physeal closure, and found residual gap and number of reduction attempts did not predict early closure, but initial displacement did. The literature on the value of open reduction and removal of interposed periosteum to lower the incidence of premature physeal closure is conflicted in these fractures, and it is likely that multiple variables are involved (energy of initial injury, amount of displacement, number of reduction attempts, age of patient).

Incomplete reduction is frequently caused by interposition of soft tissue between the fracture fragments. Grace⁷⁵ reported three patients in whom the interposed soft tissue included the neurovascular bundle, resulting in circulatory embarrassment when closed reduction was attempted. In this situation, open reduction and extraction of the soft tissue obviously is required. As noted above, a less definitive indication for open reduction is interposition of the periosteum, which causes physeal widening with no or minimal angulation. Good results have been reported after open reduction and extraction of the periosteal flap (Fig. 32-33).¹¹⁰ It is not clear that failure to extract the periosteum in such cases results in physeal arrest sufficient to warrant operative treatment. Wattenbarger et al.²¹¹ and Phieffer et al.¹⁶³ have attempted to determine the relationship between physeal bar formation and interposed periosteum, although at this time it is unclear if the periosteal flap increases the risk of physeal arrest.

Because of risk of iatrogenic damage to the distal tibial physis during closed reduction, many authors recommend the use of general anesthesia with adequate muscle relaxation for children with Salter–Harris type II distal tibial fractures. However, no study has compared the frequency of growth abnormalities in patients with these fractures reduced under sedation and local analgesia to those with fractures reduced with the use of general anesthesia. One of the authors (KS) uses general anesthesia, and an arthroscopic ankle distractor to distract the fracture before reduction, with the theoretical advantage of reducing the risk of physeal damage during the reduction maneuver (Fig. 32-34).

When closed reductions are not performed under general anesthesia, they are usually done under IV sedation. Alioto et al. demonstrated significantly improved pain relief with hematoma block for ankle fractures in a study comparing patients treated with IV sedation to patients receiving hematoma block.⁵⁴ Intravenous regional anesthesia or Bier block has also been reported to be effective for pain relief in lower extremity injuries.¹¹⁴

One advantage of reduction in the operating room with general anesthesia is the ease with which percutaneous pins can be placed to maintain reduction of the fractures. It is the experience of one of the authors that Salter–Harris I and II fractures will occasionally displace after closed reduction and above-knee casting. If there is any concern about redisplacement or stability, smooth pins can be placed at that time.

Surgeons using regional block anesthesia within the first 2 to 3 days after the fracture should consider the potential for compartment syndrome. In fractures that have a higher risk of compartment syndrome, regional anesthesia, especially peripheral nerve blocks with longer-acting agents, might delay the recognition of a compartment syndrome.¹⁴⁰

Salter–Harris Type III and IV Fractures

Salter–Harris type III and IV fractures are discussed together because the mechanism of injury is the same (supination–inversion) and their treatment and prognosis are similar. Juvenile Tillaux and triplane fractures are considered separately. In the series of Spiegel et al.,¹⁹⁶ 24.1% of the fractures were Salter–Harris type III injuries and 1.4% were type IV. These injuries are usually produced by the medial corner of the talus being driven into the junction of the distal tibial articular surface and the medial malleolus. As the talus shears off the medial malleolus, the physis may also be damaged (Fig. 32-35).

Nondisplaced Salter–Harris types III and IV fractures can be treated with above-knee cast immobilization, but care must be taken to be sure that the significant intra-articular displacement is not present. Radiographs frequently underestimate the degree of intra-articular involvement and step-off of the articular surfaces. CT imaging may be necessary to fully appreciate the degree of displacement (Fig. 32-11). Follow-up radiographs and/or CT scans in the first 2 weeks may also be necessary to confirm that no displacement occurs after casting.

Salter–Harris type III fractures of the medial malleolus may have a higher risk of physeal arrest. One study suggested that the rate of physeal arrest could be reduced by the use of open reduction and internal fixation.^102,111 Luhmann et al.¹²⁵ have recently studied a series of medial malleolar fractures with growth disturbance following treatment, and recommends anatomic reduction as fractures with as little as 2 mm of step-off went on to premature physeal closure. Others have also emphasized the importance of anatomic reduction and early treatment to reduce the risk of physeal arrest.¹⁶²

Based upon principles of fracture treatment in adults, displaced intra-articular fractures are treated with as anatomical a reduction as possible. Studies in children confirming the importance of articular reduction to within 2 mm are few (66), although most recommend anatomic articular reduction in displaced fractures involving the articular surface. Failure to obtain anatomical reduction may result in articular incongruity and posttraumatic arthritis, which often becomes symptomatic 5 to 8 years after skeletal maturity.⁴⁰ The risk of growth arrest has also been linked to the adequacy of reduction, although the literature is still unclear if anatomic reduction reduces the risk of physeal arrest (Fig. 32-36).¹¹¹ Some recent series suggest early anatomic reduction is associated with a lower risk of physeal arrest.¹⁸⁶ Closed reduction may be attempted but is likely to succeed only in minimally displaced fractures. If closed reduction is obtained, it can be maintained with a cast or with percutaneous pins or screws supplemented by a cast.

If anatomical reduction cannot be obtained by closed methods, open reduction and internal fixation or mini-open arthroscopic reduction should be carried out. Lintecum and Blasier¹²⁰ described a technique of open reduction achieved through a limited exposure of the fracture with the incision centered over the fracture site combined with percutaneous cannulated screw fixation. This technique was performed on 13 patients, 8 Salter–Harris IV fractures, 4 Salter–Harris III fractures, and 1 triplane fracture. The authors reported one growth arrest at follow-up averaging 12 months. Beaty and Linton¹⁰ reported a Salter–Harris type III fracture with an intra-articular fragment (Fig. 32-37); these fractures require open reduction for inspection of the joint to ensure that no osteochondral fragments are impeding reduction. Arthroscopic evaluation of the joint may also be an option. Internal fixation devices should be inserted within the epiphysis, parallel to the physis in patients with greater than 2 years of growth remaining, and should avoid entering ankle joint (Figs. 32-21 and 32-38).

Arthroscopic-assisted fixation of fractures with intra-articular involvement have been described by several centers. Jennings et al.⁹² presented a series of five triplane and one Tillaux fractures treated with arthroscopic assistance. The outcome was excellent for fracture reduction and ankle function. Kaya et al.¹⁰³ review 10 patients with juvenile Tillaux fractures treated with arthroscopic assistance, demonstrating excellent reduction and clinical outcomes.¹⁵³ One of the primary advantages of arthroscopic fixation is that it allows for visualization of the articular surfaces, although the need for open reduction of the metaphyseal and epiphyseal regions may still require open incisions.

Options for internal fixation include smooth Kirschner wires, small fragment cortical and cancellous screws, and 4-mm cannulated screws (Fig. 32-39). Several reports^12,20,32 have advocated the use of absorbable pins for internal fixation of ankle fractures. Benz et al.¹² reported no complications or growth abnormalities after the use of absorbable pins with metal screw supplementation for fixation of five ankle fractures in patients between the ages of 5 and 13 years. In reports of the use of absorbable pins without supplemental metal fixation in adults,^19,21,68,86 complications have included displacement (14.5%), sterile fluid accumulation requiring incision and drainage (8.1%), pseudarthrosis (8%), distal tibiofibular synostosis (3.8%), and infection (1.6%). Bucholz et al.³² reported few complications in a series of fractures in adults fixed with absorbable screws made of polylactide and suggested that complications in earlier series might be related to the fact that those pins were made of polyglycolide. A report in 1993 by Böstman et al.,²⁰ however, included few complications in a series of fractures in children fixed with polyglycolide pins. A follow-up report by Rokkanen et al.,¹⁷⁵ in 1996 reported 3.6% infection and 3.7% failure of fixation.

The main advantage of absorbable pins and screws is that hardware removal is avoided. Böstman compared the cost-effectiveness of absorbable implants in 994 patients treated with absorbable implants to 1,173 patients treated with metallic implants. To be cost-effective, the hardware removal rates required were calculated to range from 19% for metacarpal fractures to 54% for trimalleolar fractures.²² At this time, the indications for absorbable pins remain unclear.

Recent studies in the adult literature suggest second-generation bioabsorbable screws have lower complication rates, and their use may be increasing.^167,194 Additional studies in adult patients using ultrasound and MRI have not detected deleterious effects on healing with newer screw designs.^78,132 Because children are typically smaller and lighter than adults, the implants used for fixation may not need to be as strong or large as those required by adult patients. This suggests that younger patients may be better candidates for these bioabsorbable implants. The presence of the physis, and the low-grade inflammation that may accompany the dissolution of these implants, however, may increase the risk of physeal arrest, and additional studies in adult and pediatric patients will be necessary to confirm the effectiveness and safety of these devices.¹⁰¹

Salter–Harris Type V Fractures

Salter–Harris type V fractures of the ankle are believed to be caused by severe axial compression and crushing of the physis (Fig. 32-40). As originally described, these injuries are not usually associated with significant displacement of the epiphysis relative to the metaphysis, which make diagnosis of acute injury impossible from plain radiographs; the diagnosis can only be made on follow-up radiographs when premature physeal closure is evident. Spiegel et al.¹⁹⁶ have designated comminuted fractures that are otherwise unclassifiable as Salter–Harris type V injuries.