Salter Fractures

In 1963 Salter and Harris described a system of categorizing fractures involving the growth plate in relationship to the epiphyseal plate, the epiphysis, and the metaphysis (Fig. 90-1) [1]. Today, these specific fractures involving the growth plate are commonly called “Salter Fractures.” Fractures of the growth plate may cause partial or complete arrest of growth, which may result in the loss of bone length and/or subsequent development of angular limb deformities and gait abnormalities. It is imperative to understand the physiology of the growth plate and the possible consequences of trauma in order to properly assess the severity of damage, give an appropriate prognosis, and provide the correct treatment.

This chapter describes the basic anatomy of the growth plate, the relative contributions of the different growth plates to overall growth, the etiology and biomechanics of Salter fractures, and some diagnostic and prognostic guidelines.

Microscopic Anatomy

The growth plate consists of fibrous, cartilaginous, and bony components (Fig. 90-1). The fibrous component surrounds the growth plate and is divided into an ossification groove, called the groove of Ranvier, and a perichondrial ring known as the ring of LaCroix. The groove of Ranvier contributes chondrocytes for growth in both the diameter and length of the growth plate [2]. The ring of LaCroix is located between the ossification groove and the periosteum of the metaphysis and provides mechanical support for the growth plate. A biomechanical study in rabbits suggested that this structure protects growth cartilage from shear forces [3].

The cartilaginous component of the growth plate is divided into reserve (or germinal), proliferative, and hypertrophic zones. The hypertrophic zone is further divided into the zones of maturation, degeneration, and provisional calcification (Fig. 90-1). It is important to remember that damage to the reserve zone is associated with destruction of the germinal cells and, therefore, carries a high risk of resulting growth abnormalities. The hypertrophic zone is the weakest part of the growth plate and is most commonly involved in Salter fractures [1].

Immediately adjacent to the cartilaginous component is the bony component of the growth plate. This is the portion of the metaphysis in which cartilage cells are transformed into bone. The major cell type of the growth plate is the chondrocyte. Cell matrix, consisting of 70% water and 30% collagen fibrils, proteoglycans, and other noncollagenous proteins, is found as well [4]. Collagen fibers provide tension and shear resistance to the cartilage. Proteoglycans and water give resistance to pressure [5].

Vascular Supply

An intact vascular supply is necessary for cell proliferation and cartilage resorption and calcification, which are all necessary for growth, and for healing of fractures. Injury that results in disruption of or changes in the vascular networks may cause abnormal growth or cessation of growth [5]. The arterial blood supply to the growth plate consists of branches of the vascular supply to the epiphysis and metaphysis (Fig. 90-2). The multiple branches of the epiphyseal arteries arborize into the growth plate, providing vascularization to the first 4 to 10 cell columns of the proliferative zone. No vessels penetrate beyond the proliferative zone and, therefore, the hypertrophic zone is relatively avascular. Chondrocytes in the hypertrophic zone must use anaerobic glycolysis to furnish energy. Perichondrial arteries supply the fibrous structures of the growth plate. The nutrient artery provides four fifths of the metaphyseal blood supply. Branches of the metaphyseal arteries supply the remainder. Terminal branches of these vessels end in small vascular loops or capillary tufts below the last intact row of chondrocytes of the growth plate [6]. Venous drainage of the metaphysis occurs via the large central vein of the diaphysis [7].

Figure 90-1A. Normal anatomy. 1) Articular cartilage; 2) Epiphyseal cartilage; 3) Secondary ossification center; 4) Groove of Ranvier; 5) Ring of LaCroix; 6) Periosteum; 7) Cortical bone; 8) Epiphysis; 9) Growth plate; 10) Metaphysis; 11) Reserve zone or germinal layer; 12) Zone of proliferation; 13) Zone of maturation; 14) Zone of degeneration; 15) Zone of provisional calcification; 16) Zone of hypertrophy.

Figure 90-1 B-C. Salter Fractures. Figure 1b) Salter type 1 fracture through the hypertrophic zone of the epiphyseal plate; Figure 1c) Salter type 2 fracture through the epiphyseal plate and the metaphysis.

Figure 90-1 D-E. Salter Fractures. Figure 1d) Salter type 3 fracture through the epiphyseal plate and the epiphysis; Figure 1e) Salter type 4 Fracture through the epiphysis and the metaphysis.

Figure 90-1 F-G. Salter Fractures. Figure 1f) Salter type 5 fracture: Compression fracture of the epiphyseal plate; Figure 1g) Type 6 fracture: Lateral bone bridge formation.

Figure 90-2. Blood supply of the growth plate. 1) Epiphyseal artery; 2) Perichondrial artery; 3) Growth plate; 4) Metaphyseal artery; 5) Nutrient artery.+

In general, the blood supply to the growth plate does not enter through associated ligaments to the joint. However, it has been shown that in 28% of people, the femoral capital growth plate is supplied via branches of the artery of the ligament of the femoral head (epiphysis) [8]. In contrast, no such blood supply is evident in the dog [9].

Growth Plate Closure and Contribution of Different Growth Plates to Overall Growth

It is important to be aware of the closure times and the relative contribution of the different growth plates to overall growth in order to better assess the risk for possible secondary problems after a Salter fracture. It is generally agreed that the rapid growth period of the dog is between 3 to 6 or 7 months of age [10,11]. Most dogs achieve 90 to 95% of their adult size by the end of 7 to 9 months [11]. Growth plates of giant breed dogs may not close until 15 to 18 months of age. Studies on bone growth have shown significant individual and breed variations in the time of growth plate closure. These studies also report that longitudinal growth may stop prior to radiographic evidence of growth plate closure [10-12]. Radiographic evidence of growth plate closure occurs between 4 months and 12 months of age, depending on the specific anatomic site and breed of dog [10-12]. Table 90-1 shows a summary of the reported time frames of growth plate closure in the front and hind limbs of the average dog [12,13]. Growth plates that contribute a large percentage of the total axial growth of the long bones, such as the radius, ulna, and tibia, remain open longer compared with those of smaller bones, such as the carpal or tarsal bones. Cats have similar patterns of growth plate closure. Physeal closure begins at 4 months and is usually completed at 7 to 9 months of age. However, final closure of the distal radial physis in cats can occur as late as 20 months of age [14]. Studies in dogs have been performed to evaluate the relative contribution of each epiphyseal plate to total growth [15,16]. Table 90-1 summarizes the results of these studies.

Whenever the function of the growth plate is severely impaired, anatomic deformity is likely to develop. Trauma, dietary, hormonal, and genetic etiologies are clinically important to growth deformity in the dog. The following paragraphs will focus on the effect of trauma to the growth plate.

Fractures of the Growth Plate

Prevalence

Growth plate fractures are the result of trauma. Of dogs with fractures of long bones, 50 to 55% have been reported to be younger than 1 year of age [17,18]. Among the reported cases of long-bone fractures, 30% had trauma to the growth plate, and 7% subsequently developed growth deformities. In a study of 92 dogs with growth deformity as result of trauma, 75% showed a disturbance of the ulna or radius. The tibia had 4% and the femur 8% of the deformities reported [19].

Biomechanics

Any material, including bone or a specific area of the bone such as the growth plate, can be exposed to disruptive forces such as tension, compression, torsion, shear, and bending (see chapter 126). Depending on the amount of applied stress, bone may deform and return to normal configuration when the stress is reduced. This is known as elastic deformation. If the stress exceeds the elastic limit, the deformation persists. This is known as plastic deformation. When the point of failure is surpassed, the material breaks. The growth plate of immature animals is more prone to failure than either the ligamentous structures supporting the joint or the relatively elastic bone in these young animals [20]. Therefore, it is common to see growth plate injury without distortion of the joint or fracture of the bone. Although Salter and Harris applied shear stress in their original experiments, Salter fractures can also be caused by bending, torsion, tension, or compression [1,5,18,21].

Classification Systems

Salter-Harris Classification

Salter and Harris described a classification for fractures involving the growth plate, mostly intended for use in human patients, which categorizes fractures in relation to the epiphyseal plate, the epiphysis, and the metaphysis (Fig. 90-1) [1]. This same report explains results of an experimental study in rodents, demonstrating that fractures commonly developed through the zone of hypertrophy, which is mechanically the weakest zone of the growth plate. Salter and Harris also investigated healing of the different fracture types and suggested that the higher the fracture grade, the worse the prognosis for normal growth. They also postulated that interference with blood supply to the epiphysis is associated with a poor prognosis. The Salter-Harris classification is also commonly used in veterinary medicine.

Salter and Harris describe five types of fractures (Fig. 90-1) [1,5,18,21,22]. Type I represents complete epiphyseal separation through the zone of hypertrophy. The reserve zone (germinal layer of the growth plate) is usually intact.

In a type II injury, the fracture occurs partially through the growth plate and partially through the metaphysis. Type I and type II Salter fractures represent 65.5% of growth plate fractures in small animals and, therefore, are the most common types seen [22].

Type III is an intra articular epiphyseal fracture. The fracture line is not limited to the hypertrophic zone but also includes a small area of the reserve zone. Salter Type IV fractures are also intraarticular, but the fracture line extends into the metaphysis, thus completely crossing the growth plate. Type III and IV fractures often represent condylar fractures and occur most commonly in the distal humerus [22]. Salter types III and IV represent 25.5% of all growth plate fractures in the dog [22].

Type V Salter fractures are characterized by partial or complete compression of the growth plate. This crushing injury is uncommon. It is difficult to diagnose on the basis of radiographs. It may help to compare the width of the growth plate with the contralateral side, and to repeat the radiographic study at 2-week intervals to better assess damage.

Some investigators have suggested adding a type VI fracture to the traditional Salter-Harris classification system [21,23,24]. A type VI fracture involves the peripheral region of the growth plate, the zone of Ranvier. More commonly, it results from a localized contusion or avulsion of that specific portion of the growth mechanism. Peripheral osseous bridge formation commonly occurs, leading to peripherally localized epiphysiodesis and subsequent angular deformity [21,24].

Ogden Classification

The classification proposed by the human surgeon John Ogden is one concerned with injury to the growth mechanism rather than only physeal or epiphyseal injury. It is a more detailed scheme that permits further understanding of the injury to the growth mechanism as a whole. Ogden subdivided the Salter and Harris classification in order to better predict a prognosis for growth disturbances [24]. Salter type I lesions are divided into three subtypes (1A-C), type II lesions into four subtypes (2A-D), and type III lesions into three subtypes (3A-C), type IV lesions into four subtypes (4A-D). The subdivisions are made according to the injured zones and specific fracture pattern of the growth plate (Fig. 90-3). For example, in contrast with type 1A where the fracture is primarily through the zone of hypertrophic cartilage, type 1B fractures occur through the zone of degenerating cartilage and primary spongiosa, whereas type 1C fractures are associated with injury to the germinal portion of the physis. Subdivisions for type two to four injuries are according to the size and amount of fragments, and specific location of the fracture. Ogden also introduced new lesions (types VI – IX), describing injury of the ossified or nonossified epiphyseal nucleus, the fibrous structures of the growth plate, and multiple fractures of the epiphysis. In people, this classification system has been shown to provide a better prognosis for injuries of the growth plates, compared with the Salter-Harris system [25-27]. While further description of this system is beyond the scope of this chapter, this grading scheme may help to predict more accurately a prognosis for small animals with growth plate injuries [24].

Figure 90-3. (Courtesy: Dr. Ann L. Johnson, University of Illinois). Fracture of the proximal femoral physis, showing the fracture line going through the zones of reserve, proliferation, and hypertrophy. Destruction of cells in the reserve zone and the zone of proliferation increases the risk for a poor prognosis.

Classification Based on Blood Supply

Although the Salter-Harris classification is useful for a radiographic description of the growth plate fracture, this system does not consistently correlate with the clinical and histologic findings, nor does it predict the normality or abnormality of future growth [28]. An additional classification for fractures of the epiphyseal plate has been suggested in human medicine [15]. This system is based on the integrity of epiphyseal and metaphyseal blood circulation. Radiography is used to assess displacement of the fracture fragments, and scintigraphy is used to assess the degree of vascular damage. Type A fractures are described as having the circulation intact, with no displacement of the fragments. Type A fractures are not expected to have a negative effect on subsequent growth. In type B1 fractures, there is mild displacement of the fragments but some areas are still in direct bony contact. Type B2 fractures show more severe fragment displacement. In type B1 and type B2 fractures, epiphyseal and metaphyseal vessels intermingle through vertical fissures across the epiphyseal plate. In these fractures, bony bridges may form and lead to cessation of longitudinal growth. Type C fractures are described as having complete disruption of the epiphyseal circulation. These fractures carry a poor prognosis for normal growth [15,29].

Histologic Findings in Traumatic Growth Plate Injuries in Dogs

A study of the histologic appearance of traumatic canine physeal injuries that had been radiographically classified as Salter-Harris type I or type II fractures showed that in 10 of 13 cases, the fracture disrupted cells in the proliferative zone, rather than in the hypertrophic zone (Fig. 90-4) [30]. Destruction of cells in the proliferative zone may worsen the prognosis for continued growth, in contrast to the expected favorable outcome established by the Salter-Harris classification. Other experimental studies showed that the fracture sites do not always occur at the level of the hypertrophic zone, as described by Salter and Harris, but can also involve other zones of the growth plate [31,32]. These differences may be explained by the fact that force impact occurred from only one side in the Salter and Harris experiments, whereas in vivo, combined forces may be acting, resulting in a variety of fracture patterns through the growth plate. As a result of these histologic findings, it can be suggested that the Salter-Harris classification of fractures, as described for humans and rodents, may not be fully comparable to clinical and experimental evidence in dogs.

Figure 90-4. (Courtesy: J.M. Wattenbarger, M.D., OrthoCarolina, and Helen Gruber, PhD., Department of Orthopaedic Surgery, Carolinas Medical Center, Charlotte.). Vertical septa connecting the metaphysis and epiphysis at 6 days after trauma to the epiphyseal plate of a rat. These septa develop into bone bridges between the metaphysis and epiphysis, which can eventually lead to cessation of growth and the development of angular limb deformity.

Epiphyseal Bone Bridge Formation

Physeal bone bridges connect epiphyseal and metaphyseal marrow compartments, and are thought to be responsible for cessation of localized growth and development of angular limb deformities attributed to asymmetrical restriction of longitudinal growth across the physeal plate [32-36]. Although fracture healing generally occurs without complication when the fracture is contained within the cartilage of the physis, cellular debris and the formation of vertical septa followed by physeal bone bridge formation are seen when the fracture extends to the physeal–epiphyseal border. Experimental studies found that fractures may involve all regions of the growth plate [32,37]. In one study of 20 rats, the periosteum was elevated from the proximal tibiae [37]. A fracture through the growth plate was then created. The authors reported histologic evidence of vertical septa within the growth plate extending from the epiphysis to the metaphysis as early as 6 days after injury, (Fig. 90-4), which were followed by the formation of bony bridges across the physis by days 10 and 21 in 65% and 75% of the rats, respectively. The presence of bone bridges after trauma to the epiphyseal plate has previously been described; however, it was not known that physeal bone bridge formation starts to occur so soon after injury [32-36].

Diagnosis

Most types of Salter fractures can easily be diagnosed using radiography [38]; although sometimes it is difficult to determine the amount of damage, especially with Salter-Harris type V fractures. Whenever trauma to the growth plate is suspected, serial radiographs are recommended at 1- or 2-week intervals to detect abnormalities in growth. Magnetic resonance imaging (MRI) has been shown to be superior to plain radiographs and computed tomography as a means of diagnosing injuries to the growth plate [39-41]. MRI provides accurate mapping of physeal bone bridges and associated growth abnormalities that may have already developed [39]. It has been shown that an excellent correlation exists between MRI and histologic findings [42]. Knowledge obtained from MRI studies provides an accurate diagnosis that may change the initial treatment plan. Scintigraphy provides early and accurate evidence of disturbance of epiphyseal blood flow, which, in turn, may cause growth plate injury [29]. As described above, scintigraphy may be a useful alternative means for the classification of growth plate injuries.

Prognosis

About 5 to 17% of all growth plate injuries in dogs are reported to have clinically significant sequelae [18,19,36,43]. The sequelae in altered growth following growth plate fracture depend on many factors, including which physis is injured; the type and severity of injury, including displacement; the stage of physeal maturation (i.e., the age of the animal) at the time of injury; the promptness of proper diagnosis; and the method of treatment. In general, the younger the animal, especially those younger than 6 months of age, the more serious the consequences to longitudinal bone growth [43]. The degree of retardation in bone growth following physeal injury is roughly proportional to the degree of destruction to the region or zone of the epiphyseal plate [36].

Clinical lameness as a sequala to growth plate injury occurs only if the length discrepancy between legs is more than 2 centimeters in animals under 20 kilograms or more than 3 centimeters in animals over 20 kilograms [44].

Although the Salter-Harris classification system provides a good description of epiphyseal fractures, the prognosis should always be guarded. Although the majority of patients with types I and II fractures have an excellent prognosis for normal growth, radiographically undetectable physeal damage may exist that can only be diagnosed retrospectively. Damage to the epiphyseal vessels and to the fibrous component of the physis may occur in both Salter fracture types I and II, but when early reduction and fixation are possible, uncomplicated healing is expected within 3 to 4 weeks. The prognosis for type III fractures may be good if early reduction and good reconstruction of the joint surface are achieved, and if the epiphyseal blood supply is not severely disrupted. In contrast, more severe trauma carries a guarded prognosis because malalignment and arthritis may develop. Damage to the germinal layer, the joint cartilage, and to the vascular supply in type IV fractures is common, therefore, the prognosis is guarded. Type V fractures are always associated with a poor prognosis if they happen during a period of rapid growth, because of destruction of the cells in the germinal layer [45]. Partial or complete premature closure of the growth plate and the development of bone deformities are frequently observed.

In a study of distal femoral Salter I or II injuries in 17 dogs, 82.4% showed some degree of femoral growth disturbance, and the mean extent of decreased growth was 6.7% [43]. However, clinical lameness was observed in only 3 dogs evaluated [43]. The authors suggested that compression, injury to the blood supply of the germinal layer, and fissuring through the growth plate with secondary formation of bone bridges may have possibly contributed to the high incidence of decreased femoral growth. The results of this study showed that prognosis for normal growth cannot be based solely on the Salter-Harris classification system.

It has been shown that the initial displacement, number of reduction attempts, and treatment method did not significantly affect the incidence of premature closure of the growth plate in people [46]. However, it has also been shown that improved anatomic reduction decreases the incidence of premature physeal closure [46]. For example, evidence of a residual gap following reduction is associated with a 66% incidence of premature closure, whereas with no gap the incidence decreases to 17% [46]. Residual gaps may be due to entrapment of periosteum; therefore, open reduction to remove entrapped periosteum and thus prevent development of length discrepancies or angular limb deformation has been suggested [46,47].

As a general guideline, patients with injuries to the growth plates should be reevaluated at 2-week intervals for least for 6 to 8 weeks following treatment or until complete closure of the growth plates occurs. This will ensure early detection and correction of abnormalities [5].

Summary

Salter fractures involve the growth plate and may result in cessation of bone growth or development of angular limb deformities. This may lead to impaired joint function and abnormal gait. Salter and Harris developed a classification system based on radiographic appearance in order to better describe growth-plate fractures and better predict a prognosis. This radiographic description of growth-plate fractures is well accepted in veterinary medicine. However, it has been shown that the prognosis does not always correlate well with the Salter fracture type. Although it was believed in the past that the germinal layer and the zone of proliferation are not affected in types I or II Salter fractures, histologic studies have shown that injury to these zones can be found even in these “low grade” fracture types. Therefore, the initial prognosis for any Salter fracture should be guarded, and frequent reevaluation until healing and/or complete growth plate closure is recommended. Advanced imaging techniques such as MRI and scintigraphy may become useful diagnostic and prognostic tools for managing animals with Salter fractures.