14.2 Physeal injuries

The primary role of the physis is coordinated, longitudinal growth. An injury to the physis may recover without adverse effect, or it may go on to cause a growth disturbance. This may take the form of a slow-down in the rate of growth or an angular deformity, although an increase in the rate of growth is also seen after some injuries. Physeal fracture is the most common, but not the only cause of a physeal injury; many authors use these terms interchangeably. Although progress has been made in understanding the nature and types of physeal injuries, much remains unknown.

The anatomy of the physis and its relation to function is largely understood; however, less is known of the physiology of the physis and its response to injury. It is known to respond to the genetic, hormonal, and nutritional environment as well as to the local influences of blood supply, innervation, and mechanical stress. Thus the physis is not just an area of supportive tissue like the rest of the growing skeleton, but a responsive, specialized organ with a function (growth) that is subject to a range of pathological ‘injuries’ including endocrinopathy, ischaemia, autoimmune disease, and of course trauma. The various causes of physeal injury are listed in Table 14.2.1. Meningococcal disease is a particularly potent cause of multiple severe growth arrests in survivors. The physis is characteristically resistant to infiltration by malignant tumours, but some benign tumours are associated with disturbance of physeal function with deformity.

The rates of growth of the different physes vary (Figure 14.2.1), fast in some (the distal femur for example) and slow in others. Histologically, the physis is a layered structure. In humans this layered appearance is only easily seen in fast growing physes. Different authors variously subdivide them into three, four, or five zones. The main zones are the germinal layer (or resting zone) from which the cellular columns originate and which is critical to physeal function; the proliferative zone, the hypertrophic zone, and the zone of calcification (Figure 14.2.2) (see also Chapter 1.3).

The physis is surrounded by the perichondrial ring which is contiguous with the articular cartilage of the epiphysis and to which the metaphyseal periosteum is firmly attached. It encircles the physis just as periosteum encircles the bone. As well as contributing cells to the germinal layer for latitudinal growth (the zone of Ranvier), the perichondrial ring plays an important role in maintaining the structural integrity of the physis, in particular its resistance to shear and tensile forces. Logically a fracture line should take the path of least resistance and traverse the weakest of the layers, the hypertrophic zone, which is characterized by large voids occupied by apoptotic cells and relatively little supporting matrix. However, experimental studies suggest that in practice the fracture line propagates variably through the physis and can involve all layers.

Although the physis itself is devoid of vascular channels, it receives a blood supply via three separate sources (see Figure 14.2.2). Arterioles descending from the epiphyseal vessels supply the dividing cells of the germinal layer of the physis, periosteal vessels nourish the perichondrial ring and the peripheral part of the physis, whilst the nutrient artery via its branches supplies the majority of the metaphyseal side of the physis. Damage to the epiphyseal vessels, which directly affects the nutrition of the germinal layer, is usually permanent and the consequences are severe. The damage prevents cell division in the region supplied by the vessel leading to differential physeal growth. Angular and/or longitudinal growth deformity may result. The central area of the physis appears more susceptible to vascular damage than the peripheral portion. By contrast, disruption of the metaphyseal circulation is usually transient. It does not affect chondrogenesis in the germinal cell layer or its maturation, but it does temporarily hinder the subsequent transformation of cartilage to bone.

Little data is available on the incidence of physeal injuries in general. As regards fracture, if the data from three large recent studies are amalgamated, physeal fractures account for 24.8% of childhood fractures.

The much quoted Olmsted County study, which used a meticulously documented database and a stable population, reported an incidence of 279.2 acute physeal fractures per 100 000 person-years as well as a 2:1 male to female ratio and a peak incidence that corresponded to pubescence (age 14 in boys, 11–12 in girls). The latter might have a hormonal basis which results in a relative weakness of the physis at this time, although the frequency of fracture-prone behaviour at these ages may also be a factor.

Bone fracture is usually followed by pain and loss of function and the same features are apparent with physeal fractures. In a young child it may be difficult to localize the site of the injury and therefore careful examination is essential. Swelling, tenderness, abnormal movement, and crepitus are the cardinal signs of an acute injury. Do not forget to consider non-accidental injury as a possible aetiological factor, particularly in the case of fractures involving the entire distal humeral physis.

The gradual development of limb length discrepancy with or without angular deformity may be the late presenting features of trauma that is occult or has been overlooked. Other causes of physeal injury aside from trauma also typically present in this way (see Table 14.2.1).

Plain radiographs are the basic means of investigation of physeal injuries and the accurate interpretation of radiographic features is required if acute physeal fractures are to be classified. However, the physis itself is radiolucent and assessment of the injury involves looking at the adjacent radiodense epiphysis and metaphysis. A single radiograph provides only a one-dimensional view of the complex three-dimensional structure that comprises the physis and even with two views at right angles to each other, only a limited appreciation of the injury is achieved. Radiographs should be taken at right angles to the physis (rather than the diaphysis). Supplementary oblique views are often necessary when assessing physeal fractures around the knee and ankle, and stress views can be useful in injures around a uniplanar joint. For assessing late deformity following physeal injury, plain radiographs are crucial for determining the location, extent, and progression of the problem. Standing long leg films, with the patellae pointing forwards and the pelvis level (by placing the shorter leg on blocks if necessary), require a similar radiation dose to a computerized tomography (CT) scanogram and have the advantage of imaging the weight-bearing position. The advent of digital imaging and viewing has made these films far more manageable for measurement purposes.

CT is useful to identify the degree of fragmentation and the orientation of the pieces, particularly in complex fracture patterns such as Tillaux or triplane injuries. Reconstructed images can be particularly helpful. In late cases, it can be used to image osseous bars. Although magnetic resonance imaging (MRI) has theoretical advantages with its ability to demonstrate cartilage and soft tissue, the severity of the bone and soft tissue oedema which accompanies the injury limits its use and its exact role has not yet been determined.

In very young children with small secondary ossification centres, ultrasound is occasionally used for diagnostic purposes; similarly arthrography can be useful for diagnosis and guiding treatment when the injury pattern is unclear.

Classification systems of acute physeal injuries are based on radiographic appearances. The Salter-Harris classification remains the most popular (Box 14.2.2). This defines five patterns of physeal injury. The authors felt that types I, II, and III were generally associated with a good prognosis provided the epiphyseal blood supply remained intact, because the germinal layer in theory remains largely undisturbed.

The uncommon type V was thought to involve a longitudinal compression injury to the germinal cell layer which showed no visible discontinuity of bone (‘normal’ radiographs in two planes) but which had a high potential for subsequent growth disturbance. There has been considerable controversy over the nature and incidence of type V injury. Some authors feel that the true isolated type V injury may indeed be rare but that a crushing element may be associated with other fracture patterns and that this may account for the unexpectedly bad outcomes which are known to occur, on occasion, even with type II or III fractures.

Rang extended the Salter–Harris classification to include a sixth category where there was localized injury to the perichondrial ring (Figure 14.2.3). The injury was associated with a high risk of subsequent growth disturbance and angular deformity, though whether this was due to localized vascular insult or crushing of physeal cells is unclear. Although the original description was of a direct blow to the perichondrial ring, many authors have interpreted this to include avulsion injuries of the ring and open injuries where portions of the metaphysis, physis, and epiphysis are lost or damaged. Such patterns have been described classically in association with lawnmower injuries and those that occur as a result of burns, frostbite, or extravasation injuries.

The relationship of this classification system to outcome is less predictable than was originally implied by Salter and Harris. Fracture line propagation is not as constant as was once thought and outcome is known to vary with many other factors including age, site, mechanism of injury, degree of displacement, and method of treatment as well as fracture type.

Other classification systems exist. With the shortcomings of the Salter–Harris classification in mind and based on the Olmsted County epidemiological study, Peterson has recently described a new classification system (Figure 14.2.4).

The advantages of this system are that it has an anatomical basis and depicts physeal injury as a continuum from relatively minor damage (Type I), through complete transphyseal involvement (Type III), to disruption of the physis with loss of some of the physeal cartilage (Type VI). It has an epidemiological basis as well, as the fracture types occur with decreasing frequency from Type II to Type VI. It includes two ‘new’ injury patterns not included in the Salter–Harris system; the Type I fracture (which is essentially a transverse metaphyseal fracture which extends to the physis), and the Type VI injury which occurs only in open injuries and involves loss of physeal cartilage and usually some epiphysis and metaphysis as well. There is also prognostic significance to this classification, in that from Type I to Type VI the need for immediate and late surgery goes up.

Some physeal fractures do not fit neatly into these classifications systems; examples are the Tilleaux and triplane fractures, which are multiplanar intra-articular fractures of the distal tibia occurring exclusively in patients close to skeletal maturity through a partially closed physis. For this reason they rarely cause significant growth arrest in practice, despite their anatomical similarities to physeal injuries of poor prognosis.

In acute injuries, the aim of management is to reduce pain, regain function, and preserve growth potential if possible. These goals are most likely achieved by obtaining and maintaining an anatomical reduction of all components of the injury, especially the physeal plate and the articular surface, by open or closed means. A large gap between separated physeal layers tends to fill in with fibrous tissue which has the potential to become ossified and lead to growth disturbance given time and an adequate blood supply. Although the particular methods chosen will vary according to individual fracture patterns, fixation from epiphysis to epiphysis or metaphysis to metaphysis is preferred. Transfixation of the physis itself may cause further damage, but is sometimes unavoidable. Multiple passes across the physis should be avoided, and smooth removable metallic wires are ideal. Treatment needs to be prompt as physeal fractures heal in about half the time of the metaphysis of the corresponding bone.

For the late sequelae of physeal injury the focus is on preventing worsening deformity, and restoring function usually by restoring the correct anatomical relationships between body segments (see later).

The outcome of physeal injury is dependent on a number of variables (Box 14.2.3).

Of these, only the first (the anatomical extent of the lesion and displacement) can be influenced by medical treatment, which can ‘reduce’ it to some degree. However, even then the prognosis must be guarded. Severe growth disturbance can still result from seemingly benign fracture patterns, for example in Salter–Harris Type I and II distal femoral physeal fractures. This probably reflects the magnitude of the energy inflicted on the physis during the injury, and the tendency of the straight fracture line to cut across the undulating growth plate, thus involving all layers including the germinal layer. Intra-articular fractures of the femoral or radial neck often cause severe growth disturbance even in ‘favourable’ fracture types because the epiphyseal blood supply is frequently interrupted.

The two major complications are joint incongruity and growth disturbance and the two are often interrelated. Joint incongruity can follow the malunion of displaced intra-articular physeal fractures. Such fractures are often associated with displacement at the level of the physis and thus the effects of the joint incongruity become worse with time as the associated growth arrest increases the degree of the deformity. In theory, degenerative change will occur but there is little evidence to support this in the literature.

The physeal response to injury varies. Minor trauma to the physis, such as that arising in association with a diaphyseal fracture, can cause a temporary and usually benign overgrowth. With increasing energy of the insult the cellular response in the physis shifts from hyperactivity to necrosis and the rate of growth slows. Growth disturbance of this type is called a growth arrest, which may be partial or complete.

Complete arrest is uncommon but usually results in a progressive limb length discrepancy, the effect of which varies with the age at which the arrest occurs (Figure 14.2.5). Additional problems arise if the bone affected is one of a pair, for example the radius and ulna.

Temporary cessation of longitudinal growth following trauma is marked by the appearance of a so-called Harris line in the adjacent metaphysis (Figure 14.2.6). The Harris line will appear to ‘migrate’ away from the physis on sequential radiographs once growth resumes, and it should remain parallel to it. If the Harris line and the physis appear to converge however, then a partial growth arrest is occurring.

Partial arrest is more common, and usually results from trauma. The size and location of the bony tether determine the clinical deformity, though all are associated with a degree of shortening. The age at injury and remaining growth potential define the ultimate severity. Three basic categories have been described: central, peripheral, and elongated (Figure 14.2.7). Central arrests tend to cause so-called ‘fishtail’ deformities of the articular surface (Figure 14.2.8). Peripheral arrests cause angular deformity (Figure 14.2.9). Elongated arrests commonly develop after Salter–Harris III or IV fractures of the medial malleolus or lateral humeral condyle. Partial arrests may be associated with a bony bar, but not invariably so; similarly, a radiographically open physis does not guarantee that normal growth will proceed.

Once identified, partial growth arrest may be treated in various ways and a combination of treatments may be necessary, including epiphysiodesis of the contralateral limb, ipsilateral leg lengthening, corrective osteotomy, acute shortening, and bony bar excision depending on the existing clinical problem and the predicted problem at skeletal maturity. It may be better to convert a partial arrest into a complete arrest by surgical epiphysiodesis, and deal with the resulting leg length discrepancy later as in general this is an easier option than correcting major angular deformity.

The apophysis is associated with a major muscle attachment and linked to the shaft of the bone by a physeal plate. This traction physis is subjected to tensile forces which serve to shape the bone (Box 14.2.4). In contrast to other physes, these plates are not usually perpendicular to the long axis of the bone and only in certain sites do they contribute to joint congruity (proximal ulna, periacetabular, and periglenoid areas). Injury is caused by direct trauma or by a sudden major muscular contraction which disrupts the tendon–physis–bone unit when physeal injury is more likely than tendon rupture. The physeal vascular supply is via its muscle attachments and damage is unusual. If it does occur, the germinal layer of the physis is affected leading to plate closure. Most injuries happen around the time of physiological plate closure; between the ossification and fusion of the apophysis and the peak incidence for injuries is in the adolescent years. The radiological appearances of healing apophyseal fractures can be quite alarming, characterized by florid callus which can take on the appearance of an osteogenic sarcoma. Biopsy of such a lesion is not reassuring as large numbers of mitotic osteogenic cells will be seen, hence the importance of a careful history if diagnostic confusion is to be avoided. At this age, significant growth disturbance does not occur. The major apophyses are listed in Box 14.2.5.

The outcome of these injuries is dependent on the degree of displacement of the apophysis and the muscles attached to it. It is thought that muscle strength is reduced when displacement is greater than 1cm although there is little scientific evidence to support this argument and thus open reduction and internal fixation is advocated in certain cases. A symptomatic non-union or malunion is treated similarly or by excision.

New or improved imaging modalities may identify new physeal injury patterns which may in turn lead to a new classification system that might predict prognosis more accurately than current systems and therefore guide treatment more appropriately.

Distraction osteogenesis can address shortening and angular deformity, and this technique can also be applied across the physis to distract it and encourage new bone formation. It has recently been used to treat physeal bars, with or without excision. However it tends to be reserved for older children as although considerable length can be gained, the physis tends to fuse once treatment has stopped.

Replacing physeal cartilage by transplantation of physeal tissue or iliac apophysis has been attempted with some success in the experimental and clinical settings, but the technique is hampered by a lack of donor sites. Therefore considerable work is now being directed experimentally at attempts to regenerate physeal tissue. So-called mesenchymal ‘stem’ cells harvested from periosteum and embedded in a biocompatible scaffold have been shown to correct leg length discrepancy and angular deformity in an experimental model of a physeal defect in rabbits. The effects of the biological manipulation of physeal cells with cytokines such as osteogenic protein-1 (OP-1) are also under investigation, with some favourable results.

Physeal fractures usually heal quickly. For acute injuries the aim of treatment is the prompt anatomical reduction of the injured physis in the hope that normal growth will resume after healing. Predicting whether a problem will arise is not easy, although classification systems give a guide. The ultimate outcome of a physeal injury depends on many factors besides the fracture type including the age of the patient and the rate of growth of the physis concerned, and most of these factors are not under the control of the surgeon.

Whilst physeal fractures are common, permanent physeal injury resulting from them is fortunately relatively rare except in certain subtypes (e.g. Salter–Harris IV injuries). There are many other less common causes of physeal injury apart from trauma. The pathophysiological response of the physis to injury is incompletely understood but various responses have been observed including overgrowth, slow-down, growth arrest, and angular deformity. Treatment of these problems involves a variety of techniques to attempt to restore normal anatomy. Regeneration of injured physeal tissue may be available to clinicians in the future.