CHAPTER 24 Radial Head Fracture

PART A General Considerations, Conservative Treatment, and Open Reduction and Internal Fixation

Roger P. van Riet, Francis Van Glabbeek, Bernard F. Morrey

INTRODUCTION

HISTORICAL REVIEW

Before 1933,¹⁰ the literature has been well summarized by Schwartz and Young.¹¹⁰ The first description was probably made by Paul of Aegina (ad 625-690): “The ulna and radius are sometimes fractured together and sometimes one of them only, either in the middle or at one end as at the elbow or the wrist.”² Early difficulty in making the diagnosis was encountered because of “thick muscle covering.”^33,⁹³ In 1891, Hoffa described two types of radial head fractures, displaced and undisplaced.⁵⁵ Hoffa⁵⁵ and Helferich⁵⁰ recommended resection of the radial head for late deformity.

Three to 4 weeks of immobilization,¹²² passive motion, avoidance of “operative interference,”⁴⁵ removal of the fracture fragment, and excision of the entire head for severe comminution⁵⁴ were all recommended in the early 1900s. The first descriptions of a successful osteosynthesis of the fractured radial head was by Albin Lambotte in 1909.⁷¹ Other pertinent contributions include the suggestion that surgery is not a matter of election but rather of selection.⁵⁴ Although much subsequently has been written about this fracture, the focus has changed to the more complex fractures.^16,^30,⁶⁷

BIOMECHANICS AND FUNCTION OF THE RADIAL HEAD

Force Transmission

Studies in our laboratory have shown that, under the most demanding of circumstances, up to about 90% of body weight can be demonstrated across the radial head.⁸⁴ The greatest amount of force transmission occurs with the forearm in pro-nation. This is because of a screw-home mechanism that occurs during pronation with proximal radial migration.

Stability

Traditional force-displacement studies have attributed 30% of the resistance to valgus stress to the radial head.^57,^83,⁹⁶ Our studies have shown no significant resistance to valgus stability when the medial collateral ligament is intact. On the other hand, if the medial collateral ligament is deficient, the radial head is an important secondary stabilizer in preventing the elbow from dislocating.⁸⁵ Other investigators¹¹⁵ have also shown a complementary relation of the lateral collateral ligament to the competency of the annular ligament. Therefore, the radial head may be considered a secondary stabilizer to valgus stress, and it does provide an important contribution to joint force transmission (Fig. 24-1).

FIGURE 24-1 Relative contribution of radial head and medial collateral ligament to resist valgus stress. These data define the radial head as a secondary stabilizer to resist valgus stress.

INCIDENCE OF FRACTURE

Fracture of the radial head and neck has been variously reported as 1.7% to 5.4% of all fractures.^27,^66,⁸⁶ Radial head fractures occur in about 17% to 19% of cases of elbow trauma^134,¹³⁷ and account for about 33% of elbow fractures.⁷⁷ Approximately one in three cases is associated with another injury.¹²⁷

Age and Sex

In general, about 10% to 15% of these fractures involve the neck,^86,^122,¹²⁷ usually in children in whom the physis has not closed.¹⁵ Recent demographic data¹²⁷ from our institution suggest some changes when compared with previously published reports.^5,^22,^23,^29,^30,¹³⁵ Gender ratio is approximately 1:1. The male population has more severe fracture types and more often sustain associated injuries. Compared with the literature,^22,^23,^29,³⁰ the age of patients sustaining radial head fractures has increased to a mean age of 45 years, 48 years old for women and 41 years for men.¹²⁷

MECHANISM OF FRACTURE

An axial load on the pronated forearm consistently produces a fracture of the radial head similar to that seen in clinical experience (Fig. 24-2).¹²² Odelberg-Johnsson⁸⁸ observed precisely the same effect, noting that the fracture (1) occurred with posterior subluxation of the “forearm as far as the ligaments allowed,” and (2) involved the most anterior portion of the radial head when the forearm was pronated. Because the head of the radius is eccentric to the central axis of the neck,¹¹⁸ the posterolateral aspect of the radial head comes into intimate contact with the capitellum during pronation. The common occurrence of an anterolateral fracture fragment supports this theory.^29,⁶⁶ A direct blow is another uncommon cause of radial head fracture.^29,⁴¹

FIGURE 24-2 The mechanism of injury of most radial head fractures is a fall on the outstretched hand with the elbow partially flexed and pronated. A variable amount of valgus force accounts for the associated injuries that are occasionally seen.

Amis and Miller ³ recently enhanced our understanding of this and other fractures by correlating the fracture and the angle of flexion (Fig. 24-3). As is seen experimentally, either the coronoid or radial head may be fractured with the elbow in full extension, but the radial head can be fractured at greater degrees of flexion, approaching 80 degrees of the flexion arc.

FIGURE 24-3 Laboratory data reveal that either the coronoid or the radial head may be fractured with an axial load in extension. With increasing flexion, the radial head is selectively fractured.

(Modified from Amis, A. A., and Miller, J. H.: The mechanisms of elbow fractures: An investigation using impact tests in vitro. Injury 26:163, 1998.)

ASSOCIATED INJURIES

CONCURRENT FRACTURES ABOUT THE ELBOW

Based on our assessment of 333 radial head fractures seen at the Mayo Clinic, we observed that the likelihood of associated injuries strongly correlates with the severity of the radial head fracture.¹²⁷ The incidence of associated injuries increases from 20% in nondisplaced fractures to 80% in comminuted radial head fractures. The vast majority of these injuries (90%) are fractures about the elbow, mostly articular surface lesions. Approximately 20% of these articular injuries include the distal humerus, whereas in more than 90%, the proximal ulna is involved.

Fractures or cartilage injuries of the capitellum are common ^101,¹³³ but not always appreciated.¹⁹ The associated fracture of the capitellum has been studied in some detail by Ward and Nunley.¹³² About one half of capitellar fractures were shown to have associated radial head fractures, whereas approximately 2% of radial head fractures had associated capitellar fractures. Thus, this combination is rather rare.^127,¹³²

Fracture of both the olecranon and the radial head is usually considered a variety of the Monteggia fracture and has been analyzed in detail by Scharplatz and Allgower ¹⁰⁸ and others.^44,^70,^94,¹¹²

In 15% of patients, the radial head fracture is complicated by a coronoid fracture.¹²⁷ Fractures of the coronoid have been discussed in detail by Regan and Morrey.¹⁰⁰ If the fragment is large, significant elbow instability may occur.^47,^101,¹¹²

In our series, an elbow dislocation was found in about 15% of radial head fractures,¹²⁷ as compared with approximately 10% quoted in the literature.^1,^7,^27,²⁹ Combined radial head and coronoid fractures are commonly seen in elbow dislocations. The coronoid process is involved in 80% of patients that sustain a radial head fracture as part of an elbow dislocation.

Bilateral radial head fractures are uncommon and occur in about 2% of patients.^29,¹²⁷

LIGAMENTOUS INJURY AT THE ELBOW

Some degree of ligamentous injury often occurs with radial head fracture; this association is not always fully appreciated (Fig. 24-4).^17,^49,^65,^88,¹³³ An incompetent ulnar collateral ligament is suggested by an increased valgus position.^22,^86,¹⁰⁹

FIGURE 24-4 A, Radial neck fracture that developed nonunion. B, Resection of this fracture unmasked an associated medial collateral ligament disruption with resulting symptomatic unstable elbow.

Approximately 50% of associated lesions to the elbow involve clinically significant ligamentous injuries. About 10% of patients are diagnosed with a lateral or medial collateral ligament rupture, or a combination of both.¹²⁷ These clinically relevant injuries are markedly less common than previously described.

Wagner ¹³⁰ observed 24 patients with calcification in the medial collateral ligament, and Arner and associates⁵ described a 12% incidence of ulnar collateral ligament calcification. Arvidsson and Johansson⁶ found ligament or capsular disruption by arthrography with various types of radial head fractures. Johansson⁶⁵ demonstrated positive arthrographic findings in 4% of type I, 21% of type II, and 85% of type III injuries. Using magnetic resonance imaging in patients with types II or III radial head fractures, Itamura and associates⁶³ recently found lesions in the medial collateral ligament in 54%, lateral ulnar collateral ligament in 80%, and lesions of both ligaments in 50%.

This suggests that many ligamentous lesions remained subclinical. Unless formal examinations are used, associated ligamentous lesions may be undetected immediately and may predispose patients to develop chronic symptoms. The management of ligamentous injuries is discussed separately.

OTHER INJURIES

About 10% of patients with a radial head fracture sustain associated injuries other than elbow injuries. Fractures of the hand or wrist are found in about 6%.¹²⁷ Conversely, radial head fractures are found in 6% of all scaphoid fractures.¹³⁶ Shoulder injuries are uncommon (2%) and are usually found in nondisplaced radial head fractures.¹²⁷

A ligamentous injury sustained at the distal radioulnar joint at the time of the radial head fracture ^15,^28,¹¹⁶ is diagnosed in less than 1% of acute cases¹²⁸ but is well recognized as the Essex-Lopresti injury and has been the subject of several reports.^36,^46,¹²⁴ Shortening of 5 to 10 mm can be anticipated (Fig. 24-5).⁴⁶ Open reduction and internal fixation to stabilize the proximal radius is recommended. Trousdale and associates¹²⁵ reviewed Mayo experience with 20 patients. Fifteen had radial head resection without knowledge of the wrist injury. Such late detection was treated by reconstructive procedures, with a success rate of only 14%. Proper initial diagnosis was associated with an 80% satisfactory outcome and early treatment.

FIGURE 24-5 A, Fracture of the radial head treated without surgery. B, Widening of the distal radioulnar joint was present after the fracture had healed. C, The radial head was removed, and 5 mm of proximal migration occurred during the next 2 years (D).

NEUROVASCULAR COMPLICATIONS

The uncomplicated fracture of the radial head is rarely associated with any neurovascular symptoms. Severe anterior displacement may affect the radial nerve, and a rare case of posterior interosseus nerve injury has been described in the literature.¹²¹

MUSCULAR INJURY

By definition, elbow dislocation must violate the brachialis muscle, and this factor is thought to be an important variable in the development of myositis ossificans.^75,¹²³ The significance and management of this complication are discussed later in this book.

CLASSIFICATION OF FRACTURE

The first classifications of radial head fractures were described by Speed in 1924 ¹¹⁷ and Eliason in 1925.³⁷ A year later, Cutler reported a second classification²⁹ and this presented the basis for all currently used classifications. The most commonly used classification of radial head fracture is that proposed by Mason (Fig. 24-6).⁷⁷ A fourth type, the fracture dislocation, was added by Johnston.⁶⁶ In 1962, Johansson⁶⁵ added the degree of displacement to his classification and this was later combined with the original Mason classification to form the most recent adaptation of the original Mason classification.⁵⁷

FIGURE 24-6 A to C, The Mason classification of uncomplicated radial head fractures. The exact definition of the type II fracture is often difficult to determine. Type IV is not included because it represents a complicated fracture.

MAYO MODIFICATION OF MASON CLASSIFICATION

In the previous versions of this book, Morrey added a degree of sophistication to Mason’s classification of radial head fractures by dividing them into simple and complex fractures, depending on associated lesions.^81,⁸² However, the division between simple and complex fractures can sometimes be difficult, for example, when suspected ligamentous lesions are not formally investigated due to the limited clinical implications.

Therefore, based on our clinical experience of more than 333 cases, we propose to add a suffix to the original fracture type in order to quantify associated lesions about the elbow.¹²⁹

A suffix m is used if a medial collateral ligament injury is suspected or proven, but this has questionable impact on elbow stability. A capital M is used if there is an impact on stability, enough to warrant treatment. For lateral ligament injuries, l and L is used respectively. The same is done to document associated fractures to the ulna (U, u) or humerus (H, h). The suffix P is used to indicate that some sort of procedure was performed (Fig. 24-7); x for excision and F for ORIF.

FIGURE 24-7 A, Mayo classification of radial head fracture, which allows description of associated injuries. B, Hence, this type II fracture of the radial head with dislocation is termed II m,e. C, After treatment with open reduction and internal fixation and LCL repair it is termed II m LF.

FRACTURE MANAGEMENT

In general, the treatment of radial head fractures is based on the fracture type and the presence of any associated injury. These injuries involve the ligaments or articular elements with variable implications to prognosis and management (Fig. 24-8). Associated injuries should be treated on their own merit, and the following discussion will be limited to the treatment of the radial head fracture as such.

FIGURE 24-8 A, Elbow stability requires articular and ligamentous integrity. B, Absence of the radial head does not cause instability if the ulnar collateral ligament and distal radioulnar joint are intact. C, Proximal migration can occur if distal ligaments are ruptured. D, Valgus laxity may be present if the ulnar ligament is violated.

Clinical findings include hemarthrosis and painful rotation of the forearm, especially with palpation of the radial head. Radiographs are not always conclusive, but an elbow fat pad effusion in an otherwise normal radiograph is indicative of an occult radial head fracture in almost 90% of patients ⁸⁹ (Fig. 24-9). Computed tomography scanning may give additional information on the morphology of the fracture^57,¹²⁶ and may aid in planning and assessing the feasibility of a surgical reconstruction (Fig. 24-10).

FIGURE 24-9 The presence of a fat pad sign as seen here suggests an articular fracture and should be further assessed by a computed tomography study.

FIGURE 24-10 The computed tomography scan is extremely helpful to diagnose subtle associated articular injuries.

CONSERVATIVE TREATMENT

Conservative treatment of nondisplaced ⁵³ or minimally displaced radial head fractures⁵² has been shown to yield good results at a long-term follow-up of 21 and 19 years, respectively. Results of conservative treatment of displaced radial head fractures are less favorable, and patients have increased pain and decreased strength when compared with surgically managed patients.^68,¹²⁰

Sedation,¹¹¹ mobilization,^* or immobilization with^78,^98,¹¹⁴ or without¹¹¹ a plaster cast have been advocated as conservative treatment options. Minor soft tissue injuries rarely need to be addressed.^65,^78,¹³⁵

There is little question that the type I fracture (Fig. 24-11), because of its favorable prognosis and lack of concurrent soft tissue or other osseous injury, should be managed with early motion. Mason and Schutkin,⁷⁶ reporting a military experience, found a mean period of disability of 4 weeks in 18 patients treated with early motion, compared with 7 weeks in seven individuals treated with 3 weeks of immobilization. Immediate mobilization is recommended, but a delay of up to 5 days has no functional implication after a 4-week follow-up.⁷⁴

FIGURE 24-11 A, Type I fracture involving approximately 50% of the head but with less than 2 mm of displacement. B, Minimally angulated neck fracture is also considered a type I fracture.

The major residuum is loss of extension rather than pain.²⁹ Mason⁷⁷ reported that about one third of his 62 patients with this fracture lost an average of 7 degrees of extension. This may be associated with hemarthrosis of the joint. In a cadaveric study, McGuigan and Bookout ⁸⁰ demonstrated a decrease in the flexion arc of 2 degrees per milliliter fluid injected into the elbow joint. To facilitate immediate motion, aspiration of the joint, is therefore recommended^32,^34,^40,^56,⁹⁷ and pain relief could be increased by infiltrating some local anesthetic into the joint.²⁷

The most frequent complication of conservative treatment of nondisplaced radial head fractures is degeneration of the articular surfaces, which is found at long-term follow-up in about 80% of patients.^52,⁵³ Other complications include displacement or nonunion¹⁰² (Fig. 24-12). Delayed excision of the entire radial head can be considered only if any of these complications become symptomatic (Fig. 24-13).^26,^52,^53,¹⁰² Among 21 patients treated with delayed excision at the Mayo Clinic, Broberg and Morrey¹³ reported 75% with decreased pain and 77% with improved motion. The time to delayed excision ranged from 1 month to 20 years (Fig. 24-14).

FIGURE 24-12 This nonunion was asymptomatic and left alone.

FIGURE 24-13 A, Type I undisplaced fracture of the left nondominant extremity in a 49-year-old man. Treatment by early motion as tolerated resulted in displacement of the fragment (B), which progressed to a nonunion (C). D, Treatment by late excision of the radial head 4 months after the fracture decreased the pain and improved pronation and supination 6 months after the procedure.

(Courtesy of E. T. O’Brien, San Antonio, Texas.)

FIGURE 24-14 A, Type II radial head fracture was treated nonoperatively. B, Persistent pain and limitation of motion resulted in a delayed excision of the radial head. The patient had a satisfactory result.

A rare complication of nondisplaced radial head fractures treated conservatively is arthrofibrosis, which reportedly can be managed successfully with arthroscopic débridement.^18,⁷²

RESECTION OF THE RADIAL HEAD

The authors consider this issue somewhat controversial and unresolved. In general, resection is reserved for uncomplicated type III fractures (Fig. 24-15). Rochwerger and colleagues¹⁰⁴ compared the treatment of 22 type II fractures. With mean surveillance of 5 years (2 to 23 years), osteosynthesis was found to be superior to resection, as the latter had satisfactory results in just more than 50% of cases. Similar results were found in 28 patients with type III, comminuted radial head fractures. Using a meticulous surgical technique, Ikeda et al^60,⁶¹ found improved results in the group treated with open reduction and internal fixation (ORIF) over the group of patients treated with resection of the radial head.

FIGURE 24-15 A, Type III fracture demonstrating comminution. B, With long-term follow-up of 10 years, there is osteophyte formation at the ulnohumeral joint after radial head excision, but symptoms are mild. (C) The patient had no wrist symptoms.

Yet, there has been a surprising resurgence of interest in the earlier treatment of choice, which is simple excision.

Janssen et al ⁶⁴ reported 20 of 21 excellent or good results between 16 and 30 years after excision for comminuted radial head fractures. These investigators did exclude known dislocations. On the other hand, Sanchez-Sotelo et al¹⁰⁶ described excision in the face of 10 dislocations, which is consistent with a Mason type IV classification. In this group, nine of 10 were considered satisfactory, similar to that of Janssen. However, they did note a 5-degree increased valgus carrying angle and early asymptomatic degenerative changes of the ulnohumeral joint with a mean of about 4.5 years surveillance (Fig. 24-16). There have been other recent reports regarding excision without radial head replacement that have yielded surprisingly consistent outcomes. A study from Scandinavia of 61 resections noted that the timing of the resection, whether acute or delayed, was much less important than the Mason classification of the lesion. Similar observations were made from Asia by Ikeda, who reported good results at 10 years after 15 patients (four with type II and 11 with type II treated by excision). They did, however, notice that pain was present in about a third of these patients^52a and, as noted above, favor fixation.^60,⁶¹ Wallenbock et al. evaluated radial head resection in 23 instances with a mean follow-up of 17 years with personal evaluation of the entire sample. They demonstrated that 22 of 27 patients had a satisfactory outcome. They observed those with resection after a type III or IV did less well than after a type II fracture. Interestingly, there was no distinction in their group and their observations whether the patients were treated by excision in the early or in the more chronic period.¹³¹ Similarly, a report from Italy by Celli of 31 fracture-dislocations noted that only 40% were satisfactory and 60% were unsatisfactory after various treatment options.²⁴ This is consistent with Herbertsson’s observations that the fracture type has greater prognostic importance than the acuteness of the treatment.

FIGURE 24-16 A comminuted radial head fracture was excised in a patient with elbow dislocation. The patient had a good result 15 years after surgery but had radiographic evidence of arthritis.

The basic treatment rationale, therefore, of simple resection for those fractures that cannot be reduced and fixed does have some merit. This was well summarized by Wallenbock and Potsch,¹³¹ who stated that resection is recommended “as long as there are no better long-term results of prosthetic substitution of the radial head.” We hope that prosthetic substitution will prove valuable and effective. However, realistically the long-term follow-up data of those treated by resection certainly justify consideration of this as a treatment option, particularly in those that do not have an associated injury.

OPEN REDUCTION AND INTERNAL FIXATION

Surgical treatment of displaced radial head fractures has evolved from excision of fracture fragments or the entire head of the radius, to several techniques of ORIF and modern types of radial head replacement. ORIF was used sporadically in the past, largely because of the perception that it had “not proved successful in anyone’s hands.”⁹⁸ The poor earlier results were probably due to an inadequate understanding of anatomy and less refined techniques for effective internal fixation.

The fracture fragment frequently has a periosteal hinge, indicating that its viability is possible.⁴⁸ The ideal fracture for fixation is a simple, large (constituting 30% of the head) fragment that involves the anterolateral margin of the head (Fig. 24-17). However, despite the bigger technical challenge of fixing smaller fragments, biomechanical studies have also shown a benefit of fixation of fragments smaller than one third of the radial head.^8,⁹ The anterolateral margin of the head does not articulate with the lesser sigmoid notch; instrumented fixation of this fragment does not result in impingement (Fig. 24-18).^20,¹¹³ If there are multiple but large fragments, open reduction can still be performed,⁴⁸ but results can be disappointing.^69,¹⁰³