14.6 Fractures and dislocations about the paediatric forearm

Growth-plate fractures comprise 10–25% of all forearm fractures in the paediatric population. When all physeal injuries are studied, the distal radius is considered the most frequently injured single growth plate comprising 17.9–29.7% of all physeal fractures. The distal ulna is less commonly injured and consists of 2.5–4.5% of all physeal fractures. Although radial physeal fractures may occur without visible injury to the ulna, the vast majority of ulnar physeal fractures are associated with a distal radius metaphyseal or growth-plate fracture. The average age of children with distal physeal fractures varies with gender; the peak age of injury in females is at 10–11 years of age while in males it is 12–13 years of age. This disparity is likely due to differences in the age of maximal mean peak growth velocity.

The most common type of growth-plate fracture encountered at the distal radius results in a Salter–Harris type II fracture. The majority of these fractures are minimally angulated and are immobilized without reduction. Gentle closed reduction under adequate anaesthesia is usually successful for displaced fractures sufficient to cause clinical deformity on examination. Treatment guidelines reflect the large capacity for remodelling at the distal radial physis. Fifty per cent apposition is acceptable but may compromise maintenance of reduction. Dorsal–volar tilt of 20 degrees and radial angulation of 15 degrees will remodel as long as 2 years of growth remain and the physis does not close prematurely. The results of treatment are uniformly good given the substantial capacity for remodelling, particularly under ten years of age.

The long-term prognosis into adulthood is also excellent with almost all patients functioning normally even with radioulnar shortening of up to 1cm or an ulnar styloid non-union. If the fracture redisplaces after 7 days, further attempts at reduction should be avoided as evidence suggest that this can result in damage to the physis and subsequent shortening and deformity. Fractures are immobilized in a long arm cast or a well-fitted short arm cast for 3–4 weeks and protected activities are instituted after cast removal. It is recommend that percutaneous pinning after closed reduction, is considered, in cases of instability and severe soft tissue injury resulting in swelling or neurovascular compromise.

Distal ulnar physeal fractures are uncommon and may be difficult to diagnose as the secondary centre of ossification does not appear radiographically prior to the age of 6 years. Ulnar physeal injuries are more likely to result in premature growth arrest than radial physeal injuries. Treatment of these fractures depends on the associated radial injury as well as the ulnar fracture. It has been reported that up to 50% of distal ulna physeal injuries result in growth arrest. However only rarely does this result in symptoms or functional disability.

Distal radius fractures represent the commonest fracture occurring in the forearm, and there is some evidence that this incidence may be increasing. The peak incidence in boys is 11–14 and in girls 8–11 years of age.

These fractures are classified by type, displacement, and associated fractures. The type is dependent on the predominant force producing the fracture.

A compression force will result in a torus fracture, whereas a bending force will result in a greenstick or complete fracture. This is important as the former is an inherently stable pattern and it has been shown that they can be safely and effectively treated by a simple ‘futura’ type wrist brace and discharged after the initial visit. Displacement in the majority of cases is dorsal displacement with apex volar angulation where the periosteum is intact dorsally. Fractures of the distal third of the radius are usually associated with an ulna metaphyseal fracture or avulsion fracture of the ulna styloid. Rarely there is no obvious associated fracture of the ulna. Such a fracture is a Galeazzi fracture. However it has been shown that in many cases such a pattern in children would be better referred to as a ‘Galeazzi equivalent’ or ‘pseudo Galeazzi’ fracture with a radius fracture associated with a fracture occurring through the ulna physis in a bone in which the ulna epiphyseal ossification centre has not yet appeared. These injuries are difficult to diagnose and are missed in up to 41% of cases. The outcome of such an injury has been shown to be poor and if recognized should be treated in a supination cast.

Distal radius fractures with an intact ulna usually follow a fall on an outstretched hand; resultant angulation may also be accompanied with rotational deformity. Apex–volar angulation (the most common deformity) is produced by forced supination and apex–dorsal angulation with forced pronation. Galeazzi fractures usually follow a fall on an outstretched hand with hyperpronation of the forearm.

Distal radius fractures have significant remodelling capacity in the sagittal plane but less so in the coronal plane. Many authors have suggested widely varying limits of acceptable deformity with good outcomes; however, it is our experience that if the arm appears clinically deformed or there is greater than 10 degrees of angulation, reduction is necessary.

Fractures are reduced with a combination of traction, angulation, and rotation of the palm in the direction of the angulation. The fracture is brought out to length, deformity exaggeration and rotation may produce end-to-end contact. If this fails then percutaneous methods using a dorsally inserted pin into the fracture site and levering the distal fragment of the fracture into position may be successful. Typically these fractures are immobilized in short arm casts moulded to produce three-point fixation as they have been shown to be as effective as long arm casts.

Relatively high rates of reangulation in distal radius fractures from 11–62.5% have been reported. Higher rates of reangulation are noted in fractures with apex–volar angulation (supination injuries) and complete fractures versus greenstick fractures. In addition, redisplacement is also commonly seen where bony apposition following reduction is less than 50% or the fracture was initially completely displaced. Although remodelling is substantial (Figure 14.6.1), it is our recommendation that completely displaced fractures should be reduced and plaster cast immobilization should be supplemented with percutaneous K-wire fixation as this reduces the risk of subsequent redisplacement requiring further treatment.

Historically, adult forearm fractures treated non-operatively had poor results from non-union or malalignment and stiffness due to lengthy immobilization required for union. In paediatric fractures, treatment is primarily non-operative due to uniformly rapid healing, the potential for remodelling of residual deformity, and the inherent stability of the fractures once reduced.

The radius and ulna are stabilized by the interosseous membrane and by the triangular fibrocartilage complex distally and the annular ligament proximally. The pronator quadratus (distally) and pronator teres (inserting on the mid radius) actively pronate the forearm while the biceps and supinator (proximal insertions) provide supination. The insertion of these four muscles can partially account for fragment position in complete fractures. In complete distal third fractures, the proximal fragment will be in neutral to slight supination while the weight of the hand combined with the pronator quadratus tends to pronate the distal fragment. In complete proximal third fractures the proximal fragment is usually supinated.

The distal radial and ulnar growth plates are responsible for 75% and 81% of the longitudinal growth of each respective bone. This is consistent with the often made observation that distal forearm fractures have greater potential for remodelling than do more proximal fractures. Additional remodelling can also be attributed to elevation of the thick osteogenic periosteum after fracture. This periosteal sleeve aids remodelling of residual diaphyseal deformity.

An understanding of the forces leading to forearm fracture is important as reductions are often performed opposite the direction of initial injury. Paediatric forearm fractures typically follow indirect trauma, such as a fall on an outstretched hand. Direct trauma may additionally account for open fractures, severely displaced fractures, and those seen in the proximal forearm. While the final degree of fragment displacement following indirect trauma varies between greenstick and complete fractures, the initial mechanism of injury is usually the same.

In cases where sufficient force does not completely displace the fracture, an incomplete or greenstick fracture results. A greenstick fracture in one bone may accompany a complete fracture in the other. Radiographically, greenstick fractures demonstrate two-dimensional angulation which in reality is a rotational displacement. Fractures with apex–volar angulation result from an axial force applied with the forearm in supination, those fractures with apex–dorsal angulation result from an axial force applied in pronation. Reducing a greenstick fracture usually requires rotation opposite the direction of the deforming force.

When significant indirect or direct trauma exceeds the resistance of the forearm, complete fractures of both bones will follow. When completely broken by either indirect or direct forces the bones shorten, angulate, and rotate within the confines of the surrounding periosteum, interosseous membrane, and muscle attachments.

The diagnosis of forearm fractures is usually self-evident based upon history and obvious deformity. Radiographic evaluation should include anteroposterior and lateral views of the forearm. If the elbow and wrist are not adequately visualized, corresponding views are obtained to rule out radial head dislocation, supracondylar fractures and distal radioulnar joint injury.

Malrotation in complete fractures can be difficult to detect and assess. Malrotation is suspected when cortical, medullary, or bone diameters of adjacent fragments are not equal. Alternatively, malrotation can be gauged from deviations of normal orientation of proximal and distal bony prominences on radiographic analysis.

On standard anteroposterior views, the radial tuberosity is seen in profile on the medial side, while the radial styloid and thumb are seen on the opposite side. On the same view, the ulnar styloid and coronoid process are not seen. On standard lateral views the ulnar styloid is seen pointing posterior and the coronoid process pointing directly anterior; the aforementioned radial prominences will not be seen. Another useful method for determining rotation of the proximal fragment utilizes the ‘tuberosity view’. This technique uses a calibration chart to determine the rotation of the proximal radius. The distal fragment can then be manipulated and rotated into a corresponding position.

It is uniformly understood that post-traumatic paediatric angular deformities have a variable remodelling potential; however, it has not been consistently proven that rotational malalignment will also remodel. Many studies have documented better radiographic remodelling in fractures that are distal and in patients less than 9 or 10 years of age. It is important to realize that fracture location and age may not be independent variables.

It is unclear from clinical studies how much malalignment can be accepted. Studies in cadavers have demonstrated clinically significant loss of forearm rotation with residual deformities in the midshaft of the radius and ulna angulation of greater than 10 degrees.

Although several authors have demonstrated decreased remodelling potential in proximal fractures ‘marked loss of function’ is reported infrequently. Some authors have demonstrated little functional loss with decreases of forearm rotation of 35–40 degrees. It must, however, be remembered that loss of pronation can be compensated for by shoulder abduction but supination loss is not well tolerated.

The literature regarding acceptable limits of alignment is confusing and contradictory. It is our view that given the unpredictable nature of remodelling, the unclear relationship between deformity and loss of motion and the difficulty in the assessment of rotation on radiographs that the acceptable limits for alignment in both bone forearm fractures are thus. In fractures at any level in children less then 9 years of age we will accept complete displacement and 15 degrees of angulation. In children older then 9 years of age we will accept angulation in proximal fractures of 10 degrees while more distally it remains 15 degrees. However once a child approaches skeletal maturity no angulation should be considered acceptable.

Historically, incomplete fractures were treated by completing the fracture and then manipulating the bones into an acceptable position. This approach has the theoretical advantage of increasing the size of the fracture callus and decreasing the risk of refracture. Currently, it is recognized that residual angulation is a result of malrotation and should be reduced with rotation opposite the deforming force. Traction and manipulation of the apex while rotating will often assist in the reduction. The majority of greenstick fractures are supination injuries with apex–volar angulation, and these are reduced with variable degrees of pronation. It can be difficult to remember whether to pronate or supinate the hand based on the direction of the angulation. Most fractures reduce by rotating the palm towards the direction of the deformity. Those fractures with apex–volar angulation are a result of axial load in supination, therefore rotate the palm volarly (pronation). Fractures with apex–dorsal angulations are a result of pronation force, therefore rotate the palm dorsally (supination). It is not uncommon to see a greenstick fracture of one bone and a complete fracture of the other. In these cases we use the same principles of reduction by rotation. After reduction the forearm should be immobilized in the same position that reduced the fracture. Studies have documented redisplacement of 10–16% in greenstick fractures that were not adequately rotated in the cast.

Complete both-bone forearm fractures are reduced with a combination of sustained traction and manipulation. The fingers are taped to prevent sores and placed in fingertraps with the elbow at 90 degrees of flexion, counter traction is provided with 4.5–6.75kg (10–15lb) suspended from a sling over the distal humerus. The fracture and soft tissues are slowly brought out to length for 10–15min, the arm is allowed to find its own rotation. End-to-end apposition is then attempted with deformity exaggeration and direct manipulation. If attempts to achieve bony apposition are unsuccessful, complete over-riding of fracture fragments is accepted as long as rotation and angulation is within guidelines (Figure 14.6.2).

Fracture alignment in traction is assessed with fluoroscopy or plain radiographs; if adequate, the distal part of the long arm cast is applied and moulded while still in traction. Residual malrotation is addressed prior to cast application by rotating the forearm. Because most displaced both-bone fractures are in the middle region, the hand is placed in a neutral or slightly supinated position; this usually accommodates rotation and angulation. This approach has been supported in other studies that demonstrate little correlation between position of immobilization and end result. Pronation is rarely employed for complete fractures and may result in a functional loss of supination due to soft tissue contracture. Recent studies suggest that immobilizing the forearm in a cast with the elbow extended may result in a lower rate of redisplacement.

Meticulous casting is critical as several studies have documented reangulation in approximately 8–14% of cases. Some have blamed poor cast technique while others have attributed this to residual rotational malalignment. Forearm anteroposterior and lateral radiographs are taken after reduction and immobilization, minimal improvements of residual angulation can be corrected by wedging the cast.

After an adequate reduction and immobilization, patients typically return for a follow-up radiograph at 1–2 weeks after injury. Several studies have documented reangulation during the first 2 weeks. If reangulation is documented, cast removal and rereduction is recommended. Good results of rereduction have been documented if performed within a few weeks of the initial fracture. If no reangulation is appreciated, the cast is continued for 6–8 weeks or until radiographic healing. Patients are released to all activities 3–4 months after injury.

Indications for surgical intervention in paediatric forearm fractures include:

Several different techniques are available but the two most commonly applied are open reduction and internal fixation with plates and closed intramedullary nailing of either one or both bones.

As anatomical reduction is usually not needed, we prefer closed intramedullary fixation of one or both bones. Elastic stable intramedullary nailing is now the treatment of choice using a nail 0.4 times the diameter of the medullary canal.

This is introduced into the radius through an oblique drill hole in the distal metaphysis, taking care to avoid the superficial branch of the radial nerve, and through the posterolateral part of the olecranon (Figure 14.6.3). It is considered important to pre-bend the nails so that the maximum curvature occurs at the level of the fracture. Limited open reduction may be required in order to pass the intramedullary fixation. Immobilization with supplemental plaster or fibreglass long arm casts may be required. The advantages of nailing include simplicity of the procedure, union with minimal stress shielding by hardware, easier hardware removal and a better cosmetic result. Studies have shown this procedure is successful and associated with a low rate of complications compared with plate and screw fixation.

Forearm fractures treated conservatively will rarely present with significant malreduction that precludes activities of daily living. In cases of unacceptable malunion or loss of functional forearm rotation, surgical correction can be obtained with drill osteoclasis and casting or open osteotomy and plating. Both techniques will increase motion; better results are obtained when surgical correction is performed without delay.

Refracture can occur up to 9 months after original injury. Greenstick fractures are more likely to refracture than complete fractures, possibly due to weaker union from inadequate callus formation. Refracture is associated with poor clinical outcome. In these cases operative intervention may be indicated to ensure an adequate reduction. Refractures have also followed plate removal in those cases treated with primary open reduction and internal fixation and this may be lower for nail fixation.

Synostosis between the radius and ulna is very rare, but can compromise the results in cases of fracture from very high-energy injury. It is associated with fixation of both bones through a single incision and therefore this should be avoided. Extensive bone formation that restricts pronation and supination can be surgically removed. Restoration of motion following excision is usually poor in comparison with adults.

Children’s bones are more ductile and an appropriate force applied in the proper direction may result in bending without obvious fracture. Clinically, patients present with pain, tenderness, and diminished pronation and supination. Radiographs will demonstrate slight to marked bowing which extends along the distance of the affected bone (Figure 14.6.4). If bowing is subtle, the injury may be diagnosed with comparison radiographs or a bone scan performed at least 3 days after the injury.

For children under 10 years, reduction is indicated if the angulation is greater than 20 degrees or if significant limitation of forearm rotation is present. Patients older than 10 years would benefit from reduction for deformity greater than 15 degrees or when limited forearm rotation is present. Additionally, plastic deformation must be reduced when the radial head is dislocated (Monteggia equivalent) or redislocation will occur.

Treatment consists of closed reduction under general anaesthesia. A described technique of reduction involves applying a force directly over the apex of the deformity with a sandbag or rolled towel. The force must be applied for a period of several minutes in order to achieve a lasting reduction. Pressure should not be applied over the proximal or distal epiphysis. If this method is not successful, it may be worthwhile to consider open or closed osteoclasis. Following reduction, patients are typically immobilized for 6–8 weeks. This extended period of immobilization may be required due to slightly slower healing resulting from the lack of a significant periosteal reaction.

Monteggia fracture–dislocations involve an ulna fracture combined with radial head dislocation.

Bado (1967) classified Monteggia injuries into four types:

Monteggia equivalent injuries were also described and include plastic deformation of the ulna with radial head dislocation, ulna fracture with radial neck fracture, and radial head dislocation with a radial fracture proximal to the ulna fracture.

In children the anterior (type 1) and lateral (type 3) fracture patterns predominate.

It is important to identify nerve deficits as they occur more commonly in Monteggia type injuries. Clinically evident nerve dysfunction has been documented in all major nerves in 3–24% of fractures. Posterior interosseous nerve palsy is the most commonly seen deficit in association with Monteggia injuries due to tethering and compression within the supinator muscle by the arcade of Frohse. The majority of these palsies recover spontaneously within 3–6 months. Exploration may be indicated if no recovery is identified within this time frame.

The most important factor for the successful outcome following Monteggia fractures (Figure 14.6.5) is its early diagnosis. A significant number of Monteggia patterns of fracture are missed resulting in pain and loss of function. This is particularly true with greenstick fractures of the ulna which may recoil to a position of minimal displacement that masks the significant initial energy that also produces radial head dislocation and dislocations have been shown to occur up to 3 weeks after the initial injury. Regular radiographic review is necessary for up to 3–4 weeks following such injuries.

It is essential to assess the radiohumeral integrity on all radiographic views. Regardless of radiographic projection, a line drawn down the shaft of the radius should intersect the centre of the capitellum. Deviation from this relationship indicates dislocation or subluxation and therefore merits intervention.

Most authors believe that in children less than 9 years of age Monteggia injuries should be managed with initial attempted closed reduction. Successful management requires firstly reduction of the ulna fracture, secondly the reduction of the radial head and finally the alleviation of deforming forces.

Therefore type 1(anteriorly angulated) injuries are reduced by first correcting ulna length by traction and angulation by elbow hyperflexion. Secondly, gentle direct pressure over the radial head may facilitate reduction that is heralded by a gentle snap and placing the hand in supination may stabilize it. Finally these injuries are immobilized in well-moulded long arm cast in supination at 100–110 degrees of flexion. This positioning relaxes the biceps which may be a factor in late anterior radial head subluxation.

Type 2 (posteriorly angulated) injuries are uncommon in the paediatric population and the ulna fracture is reduced with traction and volarly directed force applied at the apex with the elbow in slight flexion. The radial head is reduced by direct anterior pressure and finally these fractures are immobilized in extension following reduction of ulnar alignment and restoration of normal radiohumeral articulation to prevent redisplacement.

Type 3 (lateral angulation) are reduced with a valgus force applied to the apex of the ulna fracture and direct lateral pressure on the radial head. Once reduced, the arm is immobilized in 60–110 degrees of flexion depending on initial radial head displacement. Anterolateral dislocations are better stabilized in more flexion than posterolateral dislocations which are better stabilized in less flexion.

Type 4 injuries are extremely rare and considered unstable. Closed reduction in a manner similar to type 1 injuries may be attempted but these injuries should be considered unstable and operative stabilization is probably needed to prevent radial head instability.

Regardless of injury type, patients should return to the clinic weekly for radiographs to ensure that the radius is still located and the deformity of the ulna has not recurred resulting in redislocation of the radial head. The patient should be maintained in a plaster cast for 4–6 weeks.

Operative intervention will be required in those injuries where the ulnar reduction is unstable, or the radial head is irreducible due to ulnar malreduction or an interposed annular ligament. Type 4 injuries and Monteggia equivalent injuries with radial neck fractures are relatively unstable and require operative stabilization of both radial and ulna fractures in the majority of cases. The ulna is managed with closed reduction and intramedullary fixation using a flexible nail passed distally from the olecranon. The radial head is then checked radiographically for reduction. Open reduction and ligament reconstruction or repair is recommended when the radial head will not reduce in the face of good ulnar alignment. Following repair or reconstruction of the annular ligament, the proximal radius may be further stabilized by K-wire pinning if it continues to be unstable. Because of the risk of pin failure, transcapitellar pinning of the radius has been abandoned in favour of pinning the radius to the ulna directly. The postoperative care of Monteggia fractures is similar to that for cases that are successfully managed via closed reduction. Good clinical results can be expected in paediatric patients with Monteggia injuries diagnosed within 2 weeks of injury and treated with surgery.

Proximal radius injuries in children usually result in fracture of the radial neck. Radial head fractures have been reported in the paediatric population but are considered to be rare entities. As such, this chapter will not focus on these injuries except to point out that they are usually Salter–Harris type III or IV fractures with potentially higher rates of growth abnormalities and attendant concerns for joint incongruity. Radial neck fractures in children accounted for 5.8% and 8.5% of all elbow fractures. Radial neck fractures are rarely diagnosed prior to age 4 or 5 years as the epiphysis does not ossify until after this. Radial neck fractures typically occur at two different levels—either through the physis with or without an attached metaphyseal fragment (Salter–Harris type I and II respectively) or 3–4mm distal to the physis in the anatomic radial neck.

Type 1 fractures usually result in lateral displacement of the proximal fragment with angulation that varies in direction and magnitude. The direction of angulation and displacement of the radial fragment is always lateral in relationship to the humerus; however, its angulation relative to the radius depends on the degree of hand rotation upon impact. Type 2 fractures are posteriorly displaced and result from posterior dislocation of the elbow (Figure 14.6.6). In these cases, the radial neck may be fractured following spontaneous elbow reduction or secondary to attempted closed reduction. These fractures are exceedingly rare but are important to recognize as the radial head may be rotated 180 degrees following reduction.

The majority of authors agree that angulation less than 30 degrees in the paediatric population is acceptable and should be immobilized without attempts at closed or open reduction. Translocation may be accepted when the radial head is less than 4mm displaced.

Once a decision has been made to reduce the radial neck fracture we follow a treatment algorithm until the fracture has been adequately reduced and stabilized. Initially we attempt direct pressure over the radial head under fluoroscopy control. Direct pressure is applied to the radial head as the elbow is placed into varus while partially extended. An alternative manoeuvre involves placing the elbow in 90 degrees of flexion with the elbow fully supinated. Pressure is applied to the radial head and the arm is gently pronated to reduce the fracture. If this fails, manipulation is facilitated with the use of a percutaneously placed Steinmann pin into the head of the radius or the fracture site. The pin is then used like a crow bar to lever the radial head into an acceptable position.

If closed reductions fail then we employ the technique described by Metaizeau. This involves the insertion of an intramedullary wire from distal to proximal in the radius; it is passed into the radial head which is disimpacted and by rotating the wire it reduces the fracture. The wire is left in position to stabilize the reduction.

Only if all these methods fail would we consider open reduction. Open reduction is performed through a lateral Kocher approach and caution is directed towards avoiding extensive exposure distally on the neck of the radius. Dissection in this area may disrupt the blood supply which enters proximal to the neck resulting in delayed or non-union and risking the development of avascular necrosis, or may injure the posterior interosseous nerve as it courses through the supinator muscle. Sectioning of the annular ligament may be used to facilitate reduction; however, this structure should also be repaired. If the reduction appears unstable the fracture may be stabilized with an intramedullary wire as for the Metaizeau technique, patients should be immobilized for 3–4 weeks in a long arm cast or splint.

Diminished rotation is more commonly seen in children with more severe displacement and angulation, fracture comminution, concurrent associated elbow injury, and in older patients when compared to younger patients.

Some reports have demonstrated poorer results in patients who undergo operative intervention than in those treated via closed means. These studies imply that it is better to accept some degree of malalignment rather than accept the problems of operative treatment in order to achieve an anatomical result. However, superior results have been obtained with the use of the Metaizeau technique.

Complications include aseptic necrosis of the radial head with the majority of series reporting incidences around 5%. Radial head excision in cases of aseptic necrosis may be required due to pain and loss of motion. Occasionally excessive cubitus valgus may follow excision; however, deformity is rarely associated with significant increases in pain, functional limitations, or tardy ulnar nerve palsy. Radial head enlargement has also been commonly reported in several series and may present with pain or clicking during forearm rotation. Proximal cross-union between the radius and ulna has similarly been noted in approximately 3–4% of cases with a higher incidence in patients with associated proximal ulnar injury.