Introduction
The elbow joint is uniquely designed to allow flexion and extension between the distal humerus and the ulna. Although the radial head does articulate with the capitellum, the uniaxial motion that occurs at the elbow is largely a resultant of humeroulnar motion. This has to be considered in the context of valgus alignment of the elbow, which is the norm in most humans. The degree of valgus can vary from 7 to 15 degrees and occasionally even up to 20 degrees. The functional resultant of this arrangement of the osseous structures is the ability to place the upper limb in space and increase the 3D reach of the hand. This ability is enhanced by the forearm, which is largely an articulation between the proximal and distal radioulnar joints as well as the interosseous ligament. The osseous arrangement of the proximal ulna and radius is identical but opposite to the distal radius and ulna. The proximal ulna exhibits varus, whereas the radius tends to be more linear; conversely, the distal radius exhibits valgus, whereas the distal ulna tends to be more linear. Furthermore, the proximal third of the ulna and the distal third of the radius are largely metaphyseal. The addition of the radial bow allows us the ability to rotate the forearm through nearly 180 degrees when combined with intercarpal motion, thereby further increasing the ability to place the hand in space.
In this chapter, we will consider injuries of the elbow and forearm that can affect these functional abilities of the upper limb. Namely, the focus of this chapter will be on terrible triad injuries of the elbow, radial head fractures, Essex-Lopresti injuries of the elbow and forearm, complex fractures of the proximal ulna, fractures of both bones of the forearm, and Galeazzi fractures.
Terrible triad injuries of the elbow
Hotchkiss identified that elbow dislocations with associated radial head and coronoid fractures were inherently difficult to treat and consistently had unsatisfactory outcomes. In describing this injury, he coined the term “ terrible triad of the elbow .” Our preference has been to describe these complex injuries in a more mechanistic way as posterolateral fracture-dislocations of the elbow with associated posterolateral rotatory instability (PLRI).
At the beginning of the 20th century, treatment of complex elbow dislocations involved either closed reduction and prolonged cast immobilization, or radial head excision. Although Broberg and Morrey observed satisfactory outcomes without late instability in patients with early complete radial head excision for Mason type 3 fractures, subsequent clinical and biomechanical studies have reported high rates of persistent instability and posttraumatic osteoarthritis with radial head excision in the trauma setting. In their series of 23 patients with an elbow fracture-dislocation, Josefsson et al. noted that nearly all patients with redislocation (3 out of 4) and severe joint space narrowing (11 out of 12) had an acute radial head excision. Ring et al. reviewed 11 patients with a terrible triad injury at a minimum of 2 years and found a 100% redislocation rate with an unsatisfactory functional score if the radial head was resected. Beingessner et al., in a cadaveric study, have shown altered kinematics and increased coronal plane laxity with radial head resection, especially in the setting of compromised collateral ligaments. All authors recommended that the radial head be preserved when possible, or otherwise replaced.
An improved understanding of the upgraded role of the radial head as a primary stabilizer in posttraumatic and postsurgical situations has now led to a consensus on the management of the radial head (preservation via fixation or replacement as opposed to excision) among elbow surgeons. Over the last 20 years, our focus has shifted to assessment of the contribution of the coronoid fracture to elbow stability and its management.
Anatomy and pathoanatomy
Stability of the elbow, specifically the ulnohumeral articulation, should be considered in three planes. In the coronal plane, valgus stability is provided by the anterior bundle of the medial collateral ligament and radial head. Varus stability is provided by the lateral collateral ligament (LCL) complex and anteromedial coronoid process. The common flexor (against valgus stress) and common extensor origins (against varus stress) also serve as secondary stabilizers in this plane. In the sagittal plane, posteriorly directed forces are resisted primarily by the buttress effect of the coronoid process and anterior half of the radial head. The anterior and posterior joint capsules contribute little to static stability but may have a proprioceptive role.
Rotational stability in the axial plane is provided by the lateral ulnar collateral ligament (resists external rotation or supination of the forearm) and the medial collateral ligament, in particular its posterior bundle (resists internal rotation or pronation of the forearm). Both these structures pass from the center of rotation of the distal humerus to corresponding points on either side of the proximal ulna (sublime tubercle medially and tubercle of the supinator crest laterally).
The mechanism of injury was proposed by Osborne et al. in 1966. They hypothesized that a fall on the outstretched hand first transmitted an axial load up the forearm to the trochlear notch and coronoid process, which struck the trochlea of the humerus to fracture the coronoid. At this stage, the laterally sloping inner two-thirds of the trochlea, described as a “cam” by the authors, converted the axial force into valgus and lateral rotation, allowing the coronoid to swing on an intact biceps tendon and disengage from the trochlea. The radial head rotated posteriorly off the capitellum to tear the posterolateral capsule and strip the proximal attachment of the lateral collateral ligament complex, thereby becoming trapped behind the humerus.
O’Driscoll et al, through a cadaveric study, proposed the mechanism to involve an axial load through a flexed elbow with subsequent external rotation and valgus moments as the body internally rotated on the fixed hand. They introduced the “circle of Horii concept” to describe increasing capsuloligamentous injury and elbow instability in 3 stages from lateral to medial. Stage 1 (instability) involved the lateral collateral ligament complex and posterolateral capsule. Stage 2 (perched or incomplete dislocation) involved the anterior and posterior capsules. Stage 3 (dislocation) initially involved the posterior medial collateral ligament (stage 3a) to allow the coronoid to clear the trochlea and rest behind it, with eventual involvement of the anterior medial collateral ligament (stage 3b).
In contrast to this, Schreiber et al. studied videos of acute elbow dislocations and noted that most injuries occurred in extension and pronation with a valgus moment disrupting the anterior medial collateral ligament first. Rhyou et al. also observed a similar mechanism and medial to lateral injury pattern in their MRI-based study. They speculated that a combination of valgus and pathologic forearm external rotation allowed disengagement of the coronoid from the trochlea with resultant impaction of the radial head into the capitellum. If the coronoid process was not fully disengaged from the trochlea, a simultaneous fracture of the coronoid occurred as well. Other studies have added to this debate. Luokkala et al. found a higher incidence of medial-sided injury after posterolateral elbow dislocations, but the pattern and grade of soft-tissue injury seen on MRI was quite variable.
The lack of reproducibility in mechanism and injury pattern between studies reflects the complexity of this injury. Multiple factors relating to forces acting on the elbow (force magnitude, force direction, elbow alignment), position of the elbow at the time of injury (flexion vs. hyperextension, supination vs. pronation), and patient factors (tissue laxity) likely play a role.
Clinical presentation and assessment
The degree of elbow instability after a dislocation ranges from gross static instability seen with significant osseous injury to subtle dynamic instability, and also includes a stable joint frequently seen with simple dislocations and occasionally in fracture-dislocations. The goal of clinical and radiographic assessment is to determine where an individual elbow lies on this spectrum. Stability is defined as an elbow that maintains ulnohumeral and radiocapitellar congruity through a full arc of motion and resists subluxation when subjected to physiologic forces.
Limited clinical assessment is possible in the acute period after an elbow dislocation. Ease of reduction as well as swelling and ecchymosis over the common flexor and extensor origins can provide some insight into the severity of the dislocation and structures involved.
In particular, lateral-sided bruising is indicative of a more severe injury where the thick lateral fascia has been breached. This is unusual in comparison to medial bruising, which is almost ubiquitous in elbow dislocations.
In the subacute period, provocative tests are valuable if there is doubt regarding stability or to quantify the degree of instability. Our preferred tests are the posterolateral rotatory drawer test described by Camp et al., the gravity-assisted varus stress test, and the moving valgus stress test. The first two tests are important to assess LCL injury and understand coronoid contribution to instability ( Figs. 6.1 and 6.2 ; ). The moving valgus stress test is more useful in the chronic setting for persistent medial collateral ligament pathology.
Radiographic assessment
Pre- and postreduction orthogonal radiographs as well as a CT scan with 3D reconstructions should be performed in all elbow fracture-dislocations. On prereduction radiographs, the amount of displacement of the distal humerus from the trochlear notch is suggestive of the pattern of injury and degree of soft tissue damage. Postreduction radiographs should assess ulnohumeral and radiocapitellar congruence, as well as associated osseous injuries. Coonrad et al. described the “drop sign” on follow-up lateral radiographs to suggest persistent elbow instability ( Fig. 6.3 ). Sometimes an osteochondral fracture of the posterolateral capitellum, termed the Osborne-Cotterill lesion, is seen with posterolateral elbow dislocations and PLRI ( Fig. 6.4 ). ,
A CT scan with 3D reconstructions allows better appreciation of the osseous component of the injury ( Fig. 6.5 ). This is particularly important to study the morphology of the coronoid fracture, which has relevance to elbow stability and can often predict pattern of instability. Reagan and Morrey highlighted the importance of the coronoid to elbow stability in their classification based on a single lateral radiograph. They observed a direct correlation between instability and the size of the coronoid fragment. Since then, we have realized the importance of the coronoid as a 3D structure, where the size, shape, and location of the coronoid fracture correlate with the instability pattern. O’Driscoll et al. described this through their CT-based classification. More recently, the Wrightington classification applies a three-column model to facilitate pattern recognition of elbow fracture-dislocations and provide an algorithm for their management. A 3D understanding of osseous injury is also important for preoperative planning with regard to surgical approach and fixation technique.
Other imaging (such as MRI) and intraoperative (examination under anesthesia, arthroscopy) modalities can be used to assist in the diagnosis of subtle or chronic PLRI but are beyond the scope of this chapter.
Treatment
There is limited role for nonoperative management of posterolateral fracture-dislocations of the elbow due to persistent instability and a risk of posttraumatic osteoarthritis. A small percentage of these injuries with minimally displaced radial head and coronoid fractures can be observed with serial radiographs and clinical review to ensure stability and congruence at varying degrees of elbow flexion.
Most posterolateral fracture-dislocations of the elbow, however, require surgery with the goal of achieving a stable and congruent ulnohumeral articulation through a functional range of motion. In this section, we outline our stepwise surgical algorithm, which is very similar to the standardized surgical protocol published by Pugh et al., where damaged structures are sequentially addressed from deep to superficial as seen from the lateral window.
Surgical exposure.
No single surgical approach has been shown to be superior. With posterolateral fracture-dislocations of the elbow, our preference is to use separate lateral and, if required, medial approaches. This allows access to all areas of pathology without the risk of seroma formation or flap necrosis, which can occur with a universal posterior incision. If the radial head is intact, the supinator crest is fractured or ligament augmentation is required, a Kocher approach is preferred as it gives better access to the LCL and ulna. If the radial head is fractured, an extensor digitorum communis (EDC)-split is preferred as it provides easier access to the radial head fragments and shaft. A hybrid approach can also be used, where the Kocher interval is opened and the capsular layer is developed, and then opened anterior to the LCL.
Lateral ligament complex repair.
LCL repair is crucial for restoration of stability. Deficiency is poorly tolerated as the LCL not only provides rotatory stability during axial loading, but also resists varus stress at the elbow seen with most activities of daily living, which are performed with the arm away from the body.
With acute dislocations and fracture-dislocations, a sleeve-like avulsion off the lateral humeral condyle is most common. No specific technique (transosseous versus suture anchor) has been shown to be superior to repair this avulsion. We prefer suture anchor repair. The anchor is placed just proximal and anterior to the isometric point of the lateral ligament complex as described by Morrey et al., thereby allowing the avulsed tissue/graft to sit exactly at the isometric point. We do not believe over tensioning is an issue if a truly isometric repair is achieved.
A strong primary repair is possible in most acute and subacute cases in our experience. We consider augmentation of the repair with an internal brace or biologic graft if tissue quality is poor (older patients, shredded tissue), with Z-shaped tears, and in very large arms due to increased moment arm as well as varus stress on the repair.
Less commonly, the LCL injury is at its distal attachment with a supinator crest fracture. Lag screws or a small buttress plate along the crest work well in these cases ( Fig. 6.6 ). Surgeons should also be wary of an associated Orborne-Cotterill lesion as it contributes to rotatory instability and must be appropriately addressed ( Fig. 6.4 ).
Radial head.
Radial head excision should be avoided in the acute setting as it is associated with increased strain on injured and repaired capsuloligamentous structures leading to persistent instability as well as 100% load transmission through the ulnohumeral articulation, resulting in accelerated wear of this joint.
We consider patient age, occupation (manual work) and degree of comminution in deciding between radial head fixation and replacement. We attempt fixation in young patients involved in heavy manual work as they risk early prosthetic loosening. Our preference for fixation is the tripod technique with locking screws, and we avoid radial neck plates due to risk of stiffness unless absolutely necessary ( Fig. 6.7 ). If the radial head fracture involves more than three articular fragments, we have a low threshold to replace it as a robust lateral column is critical to protect the compromised medial side, and tenuous fixation will only lead to late failure. This is supported by Ring et al. in their retrospective case series.
From a technical standpoint, radiocapitellar overstuffing and bipolar radial heads should be avoided. Overstuffing leads to accelerated radiocapitellar and ulnohumeral wear, and paradoxically creates instability ( Fig. 6.8 ). Radial head size is best assessed intraoperatively under direct vision in relation to the lateral edge of the coronoid and proximal edge of the lesser sigmoid notch.
Coronoid.
The typical coronoid fracture pattern in a posterolateral fracture dislocation is a small anterolateral fracture involving the “tip” region ( Fig. 6.5 ). Although the traditional approach has been to spend time fixing these fragments with suture techniques, it has been our practice to not fix these fragments or suture the capsule given that this central column is not critical for stability, and stable fixation of these fragments is difficult to achieve. We have not seen any recurrent instability related to this approach by not having sutured these fragments. This clinical view is substantiated by other authors who have demonstrated no significant change in elbow stability and kinematics with suture fixation of the coronoid so long as all other derangements (radial head, lateral ligament complex and, if needed, medial collateral ligament) are appropriately addressed.
For larger anterolateral facet or bi-basal coronoid fractures, our preference is arthroscopic fixation with posterior-to-anterior screws ( Fig. 6.9 ). Even though intuitively posterior-to-anterior screws may not provide the same degree of compression as anterior-to-posterior screws, they have been shown to be biomechanically superior, possibly because the coronoid fails in shear with these injuries. We use a separate medial approach through the Hotchkiss interval for large, comminuted fractures extending into the anteromedial coronoid. Anterior-to-posterior screws or buttress plating can be used for noncomminuted fractures, whereas an extra capsular plating technique is used for comminuted ones. The excised native radial head or olecranon tip can be used as bone graft for coronoid reconstruction in situations where there is coronoid bone loss ( Fig. 6.10 ).
Medial collateral ligament.
With posterolateral fracture-dislocations of the elbow, most elbow surgeons agree that routine repair of the medial collateral ligament is not warranted. Reasons include protection of the medial side by restoration of a robust lateral column and better tolerance in patients since the medial side is only stressed in select sport-specific activities such as throwing. In our practice, we consider repair of the medial side if there is persistent ulnohumeral joint widening during extension, exacerbated by pronation of the forearm after all other pathology mentioned previously has been adequately addressed.
Salvage.
Our experience has been that salvage options such as internal or external fixators are very rarely required after robust fixation of the bony and soft-tissue stabilizers using the techniques described. We reserve such options for extreme situations such as high-energy open injuries with associated vascular injury or tissue loss.
Rehabilitation.
Immobilization is poorly tolerated by the elbow. It is therefore important to leave the operating theater with a stable elbow, which allows early active motion to reengage muscle groups. We do not advocate a plaster cast or bracing in the postoperative period as they may impart a distraction force on the ulnohumeral joint and prevent recruitment of the dynamic stabilisers. Once wound healing is confirmed, all patients are provided with a compressive sleeve bandage to promote proprioceptive feedback. For the initial 6 weeks, patients are instructed to perform exercises in the supine and overhead position, which converts the normal gravity distraction force to a compressive force. In an effort to protect the lateral repair, forearm rotation exercises are performed with the elbow at 90 degrees, while elbow range of motion is performed with the forearm pronated. Varus stress imparted by movement of the arm away from the body is also avoided during this period.
Prognosis
Advances in our understanding of posterolateral fracture-dislocations of the elbow have improved the prognosis following this condition, such that if each pathoanatomical component of the injury is addressed appropriately and rehabilitation is commenced early, the outcome is not usually as “terrible” as once thought. Poor outcomes tend to occur due to subclinical chondral injuries, persistent microinstability or failures of repair which may be multifactorial but include surgical and patient factors.
The essex-lopresti injury
The Essex-Lopresti injury is characterized by the triad of distal radioulnar joint (DRUJ) disruption, interosseous ligament (IOL) rupture, and radial head fracture. It is named after the British trauma surgeon Peter Essex-Lopresti, who reported on the injury in 1951. Although this injury only occurs in approximately 1% of all radial head fractures, up to 80% of these injuries may be missed at initial presentation. , Failure to diagnose an Essex-Lopresti injury accurately can lead to devastating sequalae affecting the entire upper limb.
Anatomy
The forearm is a link between the wrist and elbow. Under normal conditions, the forearm acts as a stable ring formed by two bones (radius and ulna) and three articulations (proximal radioulnar joint [PRUJ], IOL, and DRUJ). Each of these five components contributes to the stability of the ring, and any injury involving two or more components makes the ring unstable. In this section, we discuss forearm anatomy specifically as it relates to the Essex-Lopresti injury.
At the elbow, the radial head and annular ligament are key players in providing forearm stability. The radial head, through its contact with the capitellum, acts as a bony buttress against proximal migration and valgus drift of the radius. The annular ligament encircles the radial neck to provide transverse proximal radioulnar stability.
The IOL between the radius and ulna is the main stabilizer in the central portion of the forearm. Noda et al. divided it into proximal, middle, and distal parts. The central band in the midportion of the IOL is its strongest component, with a width of approximately 1 cm. It originates 7.7 cm from the radial head at the interosseous crest of the radius and inserts 13.7 cm from the olecranon tip into the interosseous border of the ulna. Fibers of the central band are oriented at an angle of 21 degrees to the axis of the ulna. , Histologically, the IOM and central band have features of both tendon and ligament. They are composed of highly organized collagen bundles with little elastin much like a tendon, and yet they have an origin and insertion into bone like a ligament. ,
Distal forearm stability at the DRUJ is primarily by the triangular fibrocartilage complex (TFCC), which includes the dorsal and volar distal radioulnar ligaments as well as the pronator quadratus muscle. Some stability is also provided by the distal oblique bundle of the IOL, which originates in the distal sixth of the ulnar shaft to the inferior edge of the sigmoid notch, where its fibers blend with the DRUJ capsule and distal radioulnar ligaments. ,
Biomechanics and pathophysiology
The forearm acts as an integrated unit with three important functions: load transmission, rotational movement, and muscle attachment. Longitudinal and transverse stability are critical for these functions to occur smoothly. Longitudinal stability is primarily dependent on the radial head through its contact with the capitellum, while the central band of the IOM and distal radioulnar complex act as secondary stabilizers. This is supported by Rabinowitz et al., who demonstrated up to 7 mm of proximal radial translation with radial head excision in the setting of an intact IOL and DRUJ. Transverse stability is primarily dependent on the IOL, which prevents splaying between the radius and ulna.
Under normal conditions, 80% of load transmission across the wrist occurs through the radiocarpal articulation with 20% occurring across the ulnocarpal articulation. This 80:20 ratio translates into 60% radiocapitellar and 40% ulnohumeral load transmission at the elbow. The IOL is solely responsible for this load transfer from the radius to the ulna in the forearm. Disruption of the IOL will result in separate and independent load transmission along the radius and ulna with clinical consequences at the DRUJ and radiocapitellar joint because of the increased loading. , Forearm rotation involves movement of the radius about a fixed ulna. Stability at the proximal radioulnar joint, IOL, and DRUJ are important for this to occur.
Essex-Lopresti injuries are generally associated with a high-energy mechanism such as a fall on the outstretched hand from a height. This results in an axial load propagating up the forearm with abnormal longitudinal translation of the radius relative to the ulna. The force is dissipated initially through disruption of the DRUJ and IOL, eventually exiting through the radial head as a comminuted fracture.
Clinical presentation and assessment
After an acute high-energy injury, clinical signs such as wrist pain, swelling, and DRUJ tenderness may go unnoticed or be masked by more serious distracting injuries. The signs are also very nonspecific. Mild wrist tenderness and partial IOL injury are common with most radial head fractures. These, however, do not necessarily indicate longitudinal instability or a true Essex-Lopresti lesion. , When one considers how easy these injuries are to miss and the poor outcomes associated with chronic instability, we recommend preoperative clinical and radiographic assessment and intraoperative screening of all radial head fractures ( Fig. 6.11 ).
Radiographic assessment
Initial imaging should include orthogonal plain radiographs of the elbow and wrist if Essex-Lopresti injury is clinically suspected (any radial head fracture associated with a high-energy mechanism). On elbow radiographs, degree of fragment displacement and impaction often correlate with injury complexity. Phadnis et al. described the “empty space sign” on lateral elbow radiographs to suggest longitudinal forearm instability ( Fig. 6.12 ). It is characterized by the radial shaft resting up against the capitellum with splaying of articular fragments around the shaft due to significant proximal migration.
Wrist radiographs should look for DRUJ dislocation and/or positive ulna variance ( Fig. 6.12 ). The amount of positive ulna variance to indicate longitudinal instability is, however, debatable. Duckworth et al. observed that radial shortening of up to 4 mm was not associated with forearm instability. Rabinowitz et al. noted up to 7 mm of radial migration with radial head excision in the setting of an intact IOM and DRUJ. We recommend that any change in ulna variance compared to the contralateral wrist be considered significant enough to warrant further imaging and/or intraoperative screening.
Ultrasound and MRI have also been used in the diagnosis of IOL injury but should be interpreted with caution. The normal IOL appears as a hyperechoic line between the radius and ulna on ultrasound. Disruption of this line over a 2 cm distance represents IOL rupture. Ultrasound can also be used in dynamic mode to look for the “muscular hernia sign” seen with IOL tears. MRI has similar accuracy to ultrasound in the diagnosis of IOL injury, with hemorrhage and edema in the interosseous space being the most consistent signs. , The main drawback of MRI, however, is overdiagnosis of longitudinal instability. Hausmann et al. found a high incidence of oedema associated with partial IOL tears on acute MRI even in patients with undisplaced radial head fractures.
Intraoperative assessment
If the diagnosis remains uncertain after clinical and radiologic assessment, the forearm should be assessed intraoperatively under fluoroscopy. Davidson et al. described the axial compression test to assess for longitudinal forearm instability. A test is considered to be positive if there is ≥5 mm of change in ulnar variance. The radius pull test involves a proximally directed force at the radial neck after excision of radial head fragments. Translation of >3 mm is considered to be consistent with a significant IOL injury. Soubeyrand et al. originally described the radial joystick test in forearm pronation. Subsequent to this, Kachooei et al. performed the test in extension and supination. They found >5.5 mm of lateral displacement in this position to have 100% sensitivity and 90% specificity in identifying IOL disruption.
Chronic instability
Patients with established instability generally present with longstanding wrist and lateral elbow pain. There may be DRUJ asymmetry with dorsal prominence of the ulnar head. Power grip in these patients is often weak, and there is a block to forearm rotation, especially pronation.
Assessment of chronic instability should be aimed at identifying the underlying cause (missed injury vs. suboptimal treatment), which will guide management. There may be a history of an axial injury with a malunited radial head fracture. With previous surgery, we look for early failure of fixation, early loosening of arthroplasty, or radial head excision to suggest inadequate initial management ( Figs. 6.13 and 6.14 ). The presence of capitellar changes due to proximal migration and carpal changes due to ulnocarpal abutment are also important considerations in management, but they are largely evident much later in the evolution of this injury and often portend a guarded prognosis. The role of advanced imaging and dynamic tests in the chronic setting is unclear.
Treatment
The goal of managing these complex injuries is to restore forearm stability and function and to prevent sequelae of chronic instability. In this section, we discuss management of acute and chronic longitudinal forearm instability separately as strategies at the elbow, forearm, and wrist somewhat differ.
Acute instability.
Acute treatment has traditionally focused on restoring proximal length and stability and DRUJ stabilisation. Addressing the IOL injury acutely through repair or reconstruction remains controversial.
Radial head.
Radial head options in the acute setting include reconstruction or replacement, not excision ( Fig. 6.13 ). In the absence of a radial head, 90% load is borne through the IOL which itself is compromised. Trousdale et al. have reported a more favorable outcome in patients treated acutely with radial head retention or replacement over excision.
Radial head fractures associated with Essex-Lopresti injuries tend to be significantly more displaced and comminuted ( Figs. 6.11 and 6.12 ). There is a strong correlation between degree of displacement or comminution and fixation failure. , In the setting of compromised secondary stabilizers (IOL and DRUJ), tenuous fixation must be avoided at all costs as it will not withstand increased loading through the radiocapitellar joint and lead to early failure. Surgeons should have a low threshold to replace the radial head in such complex injury patterns.
Radial head replacement is technically more challenging, with less favorable outcomes in these patients. The head may be too comminuted to piece back together, making sizing difficult. It is also easier to get the radial length wrong due to loss of normal longitudinal constraints. Comparison to contralateral ulna variance once the radial head has been addressed has been shown to be a reliable method to ensure restoration of radial length. , Another option is to pull the radius out to length based on contralateral ulna variance and temporarily pin the DRUJ in this reduced position while the radial head is implanted. The pin is removed after insertion of the prosthesis. We recommend a combination of these techniques and the use of reliable local landmarks (proximal edge of lesser sigmoid notch, lateral ulnohumeral joint space) to ensure proximal radial parameters are accurately restored. ,
There is a higher risk of capitellar wear and aseptic loosening after radial head replacement in Essex-Lopresti injuries ( Fig. 6.14 ). This is due to increased forces at the radiocapitellar joint with compromised secondary stabilizers (IOL and DRUJ). Compared to metallic bearing surfaces, there is some evidence to suggest pyrocarbon implants are better tolerated by articular cartilage and underlying subchondral bone. Silicone and bipolar radial heads should be avoided. Silicone implants do not have the material properties to resist increased axial loading and are associated with fragmentation and synovitis. Bipolar prostheses do not restore the rigid lateral column required for IOM healing.
DRUJ.
The stability of the DRUJ should be assessed through range of motion once the radial head has been addressed. In many acute cases, the DRUJ will be stable. If the DRUJ remains unstable, options include acute TFCC repair or temporary cross-pinning of the distal radius and ulna. Our preference is to repair the TFCC injury, which is most commonly a sleeve-like avulsion that can easily be reattached with a single small suture anchor. We avoid leaving pins in as they are associated with complications (infection, breakage), including risk of recurrent instability once removed if the central band of the IOM fails to heal. If pins are used at all across the distal forearm in such situations, we strongly advocate that the pins be exposed on both the radial as well as the ulnar side. By doing so, it allows the treating surgeon to retrieve any broken pins from either side in the office, without the morbidity of an additional operative procedure.
IOL.
There has been renewed interest in the IOM component of the Essex-Lopresti injury in both acute and chronic settings. This is because studies show that even after restoration of radial length, the IOL may not heal adequately due to muscle interposition and poor inherent healing capacity. This will prevent normal load transfer from the radius to the ulna and result in increased forces through the radial head and capitellum. Tejwani et al. noted better redistribution of loads between the wrist and elbow when radial head replacement was combined with IOM reconstruction compared to replacement alone.
Most surgeons prefer IOL reconstruction over repair given its poor healing potential. Various grafts (autograft and allograft) patellar tendon, palmaris longus, flexor carpi radialis, braided polyethylene synthetic materials such as the LARS ligament (Corin, Cirencester, UK), fixation methods (bone plugs, cortical buttons), and techniques (pronator teres routing) of central band reconstruction have been described. To our knowledge, no single technique has been shown to be superior to the others, and none of the grafts replicate the stiffness of the native IOL. We prefer the syndesmosis TightRope (Arthrex, Naples, FL, USA) for our reconstructions and have found it to provide adequate durability, tensile strength, and stiffness.
Chronic instability.
Management of established instability generally requires intervention at the radial head, IOL, and DRUJ ( Figs. 6.15–6.17 ). In chronic cases, proximal migration becomes fixed with loss of space previously occupied by the radial head. This makes accommodating a radial head prosthesis difficult. The other issue with chronic instability is an abnormal capitellum (capitellar osteopenia with radial head resection or capitellar erosion with radial head replacement). Although some authors have recommended radiocapitellar arthroplasty in such situations, we do not favor this option as it does not address the underlying instability and risks overloading of the radiocapitellar prosthesis. Long-term outcomes of radiocapitellar arthroplasty are also unknown. A bipolar prosthesis is a suitable option in this setting due to the contracted interosseous space and maltracking issues seen with established instability. ,
In some situations, radial head replacement may not completely restore radial length, resulting in a residual positive ulna variance. A laminar spreader placed in the radiocapitellar space can help restore length and stretch the contracted tissue. On occasion, a concomitant joint-leveling procedure, most commonly an ulna-shortening osteotomy, may be performed.
Established instability also warrants IOL reconstruction. Our technique of IOL reconstruction is the same for acute and chronic instability. As mentioned earlier, we prefer the syndesmosis TightRope (Arthrex, Naples, FL, USA) for this. Salvage options such as the Sauvé-Kapandji procedure and radioulnar synostosis (one bone forearm) are generally reserved for patients who have failed multiple previous surgeries as they are associated with poor functional outcomes.
Early diagnosis and acute treatment of Essex-Lopresti injuries provides the best opportunity for a favorable outcome. Prognosis in chronic instability despite intervention is poor.
Radial head fractures
Radial head fractures account for 0.2% of all visits to the emergency department and have an incidence ranging from 2.5 to 2.8 per 10,000 per year. They comprise approximately one-third of all fractures about the elbow and are most common in the middle-aged population. There is a slight female predominance to these injuries.
Undisplaced marginal radial head fractures (Mason type 1) are most common. Although isolated injuries are much more frequent, complex injury patterns with associated osseous and/or soft tissue lesions occur in about 30% of cases. In this section, we will focus on isolated radial head fractures.
Anatomy
The anatomy of the proximal radius is quite complex and variable both between individuals and genders. It best correlates with the contralateral side. , The axis of the radial head is offset from the axis of the neck by approximately 4 mm, which in turn is offset about 15 degrees relative to the diaphyseal axis. ,
The native radial head is elliptical with major and minor outer diameters. Proximally, it has an eccentrically placed concave articular dish that is also elliptical with its own maximum and minimum diameters. , The dish is covered with hyaline cartilage and articulates with the convex capitellum of the distal humerus. The peripheral rim of the radial head is convex. It articulates with the lesser sigmoid notch of the proximal ulna and sits in a sulcus between the lateral trochlear ridge and capitellum proximally. The articular 270 degrees of the rim is lined by thick hyaline cartilage, whereas the nonarticular portion has a thinner cartilage layer. This 90 to 110 degree “bare area” that does not articulate with the PRUJ through an arc of forearm rotation has been described as the safe zone for hardware placement. ,
Yamaguchi et al. looked at the intraosseous and extraosseous vascular anatomy about the elbow in a cadaveric study. They found the radial recurrent artery to be the major blood supply to the proximal radius. A direct branch of this vessel enters the nonarticular bare area of the radial head before bifurcating into an intraosseous subarticular plexus. At the level of the radial neck, smaller branches of the radial recurrent and interosseous recurrent arteries form an extraarticular ring from where further branches perforate the neck just distal to its capsular attachment. Once intraosseous, these vessels travel proximally to the margins of the head. An understanding of this is important as it bears relevance to risk of nonunion and avascular necrosis when dealing with comminuted fractures, as well as with soft tissue dissection and hardware placement.
Function and pathophysiology
The proximal radius plays an important role in stability and load transfer from the hand to the shoulder. The radial head provides stability in all three planes of motion. In the sagittal plane, it contributes to longitudinal stability of the forearm with the IOL, DRUJ, and TFCC and acts as an anterior buttress with the coronoid process against posteriorly directed forces. In the coronal plane, it acts as a secondary restraint to valgus stress (primary stabilizer is the anterior bundle of the medial collateral ligament). The radial head also indirectly provides varus and rotatory stability by tensioning the lateral collateral ligament (LCL) complex. This contribution to stability becomes even more critical postinjury and surgery when other stabilizing structures are compromised.
Although approximately 60% of load transmission across the elbow is through the radiocapitellar articulation, the amount and pattern vary based on the position and valgus angle of the elbow. Loading through the lateral compartment is greatest in extension and pronation. The radiocapitellar contact area is also affected by these factors.
Simplistically, radial head fractures are caused by a fall on the outstretched hand. Multiple factors relating to forces acting on the elbow (force magnitude, force direction, elbow alignment), position of the elbow at the time of injury (flexion vs. hyperextension, supination vs. pronation), and patient factors (tissue laxity), however, determine the resultant injury pattern.
Isolated fractures are generally due to a fall with the elbow in extension and pronation as this results in the greatest amount of axial load through the radiocapitellar joint. More complex injury patterns where radial head fractures are associated with ligamentous or other osseous injury tend to involve significant coronal and rotational forces in addition to the axial load.
Clinical presentation and assessment
The goal of clinical and radiographic assessment is to determine whether the radial head fracture is an isolated injury or part of a more complex injury pattern with resultant forearm or elbow instability. It also guides management.
Radial head fractures in the setting of elbow dislocation and significant soft tissue injury are often an indication of a more complex injury. Key components of clinical evaluation include assessment of block to motion and the stability of adjacent joints. This is often better appreciated in the subacute period.
Our preferred tests to assess the LCL complex and for varus instability are the posterolateral rotatory drawer test described by Camp et al. ( Fig. 6.1 and ) and the gravity-assisted varus stress test ( Fig. 6.2 ). For valgus instability, we use the valgus stress test for acute medial collateral ligament injury and the moving valgus stress test in chronic situations. Longitudinal forearm instability should be suspected in patients with associated wrist and forearm pain.
Radiographic assessment
Imaging should include orthogonal radiographs and a CT scan with 3D reconstructions. Elbow radiographs are useful to assess ulnohumeral and radiocapitellar joint congruence, and to look for associated osseous injuries seen with elbow instability (most commonly coronoid fractures). We also obtain wrist radiographs in patients with forearm pain or radiographic signs suggestive of longitudinal instability (significant proximal radial migration and/or radioulnar divergence on elbow radiographs). Particular attention on the lateral x-ray should be paid to the lateral gutter where displaced radial head fragments can be extruded to and are frequently missed. 3D CT allows better appreciation of fragment size, location, displacement, and comminution. A more complex injury pattern should be suspected with increased displacement and/or comminution.
Many radiographic classifications of radial head fractures have been proposed over the years. The most cited is the Mason classification, which divides these fractures into three types based on degree of displacement and comminution. Johnston modified the original classification by adding a fourth type to denote any radial head fracture associated with an elbow dislocation. Neither of these early classification systems, however, help to guide management. In 1997, Robert Hotchkiss introduced his modification of the Mason classification based on displacement, block to motion and comminution to better define need for surgical intervention. Despite this, there is limited consensus among surgeons on management of isolated radial head fractures.
For this reason, we tend to avoid using classifications systems in our decision making. Our clinical assessment concentrates on a block to forearm rotation, if any. Our radiographic assessment focuses on factors such as head-neck continuity (i.e., partial articular vs. complete articular fracture), presence of metaphyseal comminution, and contribution of the involved area of the radial head to stability as well as PRUJ motion.
Treatment
Our approach to management of radial head fractures depends on whether it is an isolated injury or part of a more complex injury pattern. Thorough preoperative (clinical and radiographic) and intraoperative assessments are important to determine this. Significant displacement and comminution are often associated with a more complex injury. This section will focus on the management of isolated radial head fractures. Radial head fractures in the setting of ligamentous and/or other osseous injury almost always require surgery.
There is consensus among surgeons that minimally displaced (<2 mm) radial head fractures without a block to motion can be managed nonoperatively with early range of motion exercises to minimize stiffness. Our preference is to place these patients in a sling for comfort for the first few days. They are provided with a home exercise program by an occupational therapist and undergo clinical and radiographic follow-up at 6 weeks to check motion and fracture healing. Patients should be educated about the possibility of a persistent mild (10–20 degree) flexion contracture despite early movement. Most surgeons also agree that a block to motion is an indication for surgery. Blocks are more likely with peripheral extruded fragments than central depressed fragments.
Management of isolated radial head fractures based on displacement and comminution in the absence of a block is more controversial. There is limited consensus on millimeters of displacement and/or angulation needing operative intervention, as well as the best surgical option for comminuted fractures.
Controversy based on displacement.
Akesson et al. reported a 90% good to excellent long-term outcome with nonoperative management of two-part fractures involving ≥30% of the articular surface with up to 5 mm of displacement. In contrast to this, Geel et al. recommend open reduction and internal fixation of fractures involving >25% of the articular surface and with >2 mm of displacement. Our preference is open reduction and internal fixation of isolated two- and three-part fractures involving ≥30% of the articular surface and with >4 mm of displacement, especially in young patients with good bone stock. This minimizes risk of late sequelae such as a symptomatic nonunion or malunion, painful clicking, and early radiocapitellar arthrosis. In the older patient, even displaced fractures without a block to forearm rotation can be treated nonoperatively.
Controversy based on comminution.
Evidence on whether isolated comminuted radial head fractures should be fixed, replaced, or excised is unclear. Some studies have reported favorable outcomes after open reduction and internal fixation of Mason type 3 fractures. , In contrast to this, King et al. observed a high failure rate with open reduction and internal fixation of fractures involving >3 parts. Similar findings have been reported by Ring et al. with a 93% dissatisfaction rate in fractures with >3 articular fragments. In the setting of comminution, radial head arthroplasty is also supported by two prospective randomized controlled trials comparing fixation with arthroplasty for Mason type 3 fractures. , To add to this debate, Antuna et al. noted over 90% satisfaction after radial head resection in young patients with isolated comminuted fractures at long-term follow-up. In the experience of the senior author (CSM), radial head arthroplasty is a viable and durable option for nonreconstructible fractures of the radial head in patient under the age of 40 years (Mudgal et al, 2023; unpublished data).
Our approach to isolated comminuted radial head fractures takes patient age into account with a lower threshold for fixation in young patients with good bone stock, and arthroplasty in older low-demand patients. Regardless, a stable anatomic reconstruction to allow early movement is critical if fixation is attempted. We avoid radial head resection even in uncomplicated cases as it can lead to decreased grip strength, some degree of valgus drift and proximal migration, and increased ulnohumeral loading with early arthrosis.
The rest of this section will discuss exposure and surgical options for isolated radial head fractures.
Surgical exposure.
The three most commonly described approaches to access the radial head are the Kaplan, EDC-split, and Kocher intervals. The Kaplan (between extensor carpi radialis brevis and EDC) and EDC-split (through the EDC tendon) intervals are slightly more anterior, although differentiation between them is rather theoretical in the acute trauma setting. The advantage of the anterior intervals is lower risk of iatrogenic injury to the lateral collateral ligament complex. They, however, place the posterior interosseous nerve at risk, which must be protected with forearm pronation. The Kocher interval (between extensor carpi ulnaris and anconeus) lies more posteriorly with an increased risk of inadvertent lateral ulnar collateral ligament (LUCL) injury and elbow instability. Benefits of this interval are increased distance from the posterior interosseous nerve and exposure of the LUCL and supinator crest. All three approaches can be extended proximally along the lateral epicondyle and lateral supracondylar ridge to improve exposure into the joint. A less commonly mentioned approach is the Boyd interval through a posterior incision. It involves elevation of anconeus from the proximal ulna to expose the LUCL inserting into the supinator crest. The approach is useful for radial head visualization in cases of distal LUCL injury or a supinator crest avulsion. Alternately, a supinator crest osteotomy (Wrightington approach) can be performed to expose the radial head.
Our preference is to use the EDC-split interval for isolated radial head fractures. The interval is easy to identify and stays well clear of the LCL complex. We use the Kocher interval for more complex injury patterns where competence of the LUCL needs to be assessed and reconstruction may be required.
Open reduction and internal fixation.
Radial head and neck fractures can be fixed using various types of independent screws (headless compression screws, countersunk cortical screws, locking screws) or low-profile locking plates. Choice of implant often depends on fracture configuration and surgeon preference.
We use small locking screws independently for partial articular fractures without radial neck involvement. These smaller-diameter screws provide excellent purchase even in small fragments with limited bone stock. They can also be used in a unicortical manner due to engagement of the screw head’s locking thread with subchondral bone. Results using this technique have been excellent in our case series. For more comminuted and axially unstable fractures involving the radial neck, our preference is the tripod technique with locking screws. It involves three subchondral screws passed at varying degrees of prono-supination through the articular rim and into the cortex of the distal fragment for bicortical purchase. The screws are circumferentially distributed around the radial head (tripod configuration) and provide compression through the fracture ( Fig. 6.7 ). This technique has been shown to be biomechanically superior to plate fixation and has the advantage of causing less hardware irritation. , Plates commonly cause rotational stiffness, especially in supination, and require removal in up to 70% of cases. Select fractures can also be managed arthroscopically with small case series showing satisfactory short-term results.
Arthroplasty.
Important considerations in radial head arthroplasty include implant choice and surgical technique. Many different radial head prostheses with regard to composition (metallic vs. pyrocarbon vs. silicone), shape (anatomic vs. nonanatomic), modularity (monoblock vs. modular), polarity (monopolar vs. bipolar), stability (porous press-fit vs. smooth loose-fit), fixation (cemented vs. uncemented), and stem length are now available. However, there is no clear advantage of one implant over others, and every design has its own specific set of problems. Silicone implants are no longer recommended due to a high risk of synovitis and fragmentation. Anatomic heads have the theoretical benefit of replicating native contact areas and pressures but are technically difficult to implant correctly resulting in early loosening and failure. Bipolar prostheses may be useful in the revision setting for maltracking but can cause persistent instability in acute traumatic situations. Press-fit stems are associated with aseptic osteolysis and are easier to overstuff resulting in loss of flexion. Loose-fit stems act as spacers and have an increased risk of endosteal resorption due to pistoning as well as heterotopic ossification. , Our preference is to use either a loose-fit stem (CSM, RG, AB) or a modular press-fit prosthesis (JP, MJ) due to ease of sizing.
Surgical technique is far more important for a good outcome than implant design. Radial head alignment, width, and height are critical factors to consider. The radial neck osteotomy should allow the implanted radial head to line up with the rotational axis of the forearm, which is not necessarily perpendicular to the radial neck ( Fig. 6.18 ). The minor outer diameter of the native head best correlates with intact head size ( Fig. 6.19 ). If the head is too wide, it will overload the proximal radioulnar joint and place the lateral collateral ligament complex in a biomechanically disadvantaged position. Incorrect height or overstuffing will result in decreased range of motion, especially in flexion, capitellar erosion due to increased radiocapitellar loading, and altered ulnohumeral kinematics ( Fig. 6.8 ). van Riet et al. found the proximal edge of the lesser sigmoid notch to be an accurate local landmark in assessing height. Intraoperative gapping of the lateral ulnohumeral joint has also been shown to be a reliable indicator of overlengthening. Radiologically, Athwal et al. have described a highly sensitive measurement technique based on comparison to the contralateral elbow to assess radial height. After implantation, the trial prosthesis should be screened through range of motion to ensure smooth tracking against the capitellum. In the experience and opinion of the senior author (CSM), when a surgeon is faced with a radial head size that does not exactly replicate the native radial head, it is prudent to choose a radial head size that is smaller rather than one that is larger.
